bool BaseCallerParameters::InitContextVarsFromOptArgs(OptArgs& opts){
assert(bc_files.options_set);
char default_run_id[6]; // Create a run identifier from full output directory string
ion_run_to_readname (default_run_id, (char*)bc_files.output_directory.c_str(), bc_files.output_directory.length());
context_vars.run_id = opts.GetFirstString ('-', "run-id", default_run_id);
num_threads_ = opts.GetFirstInt ('n', "num-threads", max(2*numCores(), 4));
num_bamwriter_threads_ = opts.GetFirstInt ('-', "num-threads-bamwriter", 6);
context_vars.flow_signals_type = opts.GetFirstString ('-', "flow-signals-type", "none");
context_vars.extra_trim_left = opts.GetFirstInt ('-', "extra-trim-left", 0);
context_vars.only_process_unfiltered_set = opts.GetFirstBoolean('-', "only-process-unfiltered-set", false);
// Treephaser options
context_vars.dephaser = opts.GetFirstString ('-', "dephaser", "treephaser-sse");
context_vars.keynormalizer = opts.GetFirstString ('-', "keynormalizer", "gain");
context_vars.windowSize = opts.GetFirstInt ('-', "window-size", DPTreephaser::kWindowSizeDefault_);
context_vars.skip_droop = opts.GetFirstBoolean('-', "skip-droop", true);
context_vars.skip_recal_during_norm = opts.GetFirstBoolean('-', "skip-recal-during-normalization", false);
context_vars.diagonal_state_prog = opts.GetFirstBoolean('-', "diagonal-state-prog", false);
// Not every combination of options is possible here:
if (context_vars.diagonal_state_prog and context_vars.dephaser != "treephaser-swan") {
cout << " === BaseCaller Option Incompatibility: Using dephaser treephaser-swan with diagonal state progression instead of "
<< context_vars.dephaser << endl;
context_vars.dephaser = "treephaser-swan";
}
context_vars.process_tfs = true;
context_vars.options_set = true;
return true;
};
TagTrimmerParameters MolecularTagTrimmer::ReadOpts(OptArgs& opts)
{
// Reading command line options to set tag structures
TagTrimmerParameters my_params;
my_params.min_family_size = opts.GetFirstInt ('-', "min-tag-fam-size", 3);
my_params.suppress_mol_tags = opts.GetFirstBoolean ('-', "suppress-mol-tags", false);
//my_params.cl_a_handle = opts.GetFirstString ('-', "tag-handle", "");
//my_params.handle_cutoff = opts.GetFirstInt ('-', "handle-cutoff", 2);
my_params.master_tags.prefix_mol_tag = opts.GetFirstString ('-', "prefix-mol-tag", "");
my_params.master_tags.suffix_mol_tag = opts.GetFirstString ('-', "suffix-mol-tag", "");
ValidateTagString(my_params.master_tags.prefix_mol_tag);
ValidateTagString(my_params.master_tags.suffix_mol_tag);
// Overload to disable molecular tagging
if (my_params.min_family_size == 0)
my_params.suppress_mol_tags = true;
else if (my_params.min_family_size < 1) {
cerr << "MolecularTagTrimmer Error: min-tag-fam-size must be at least 1. " << endl;
exit(EXIT_FAILURE);
}
my_params.command_line_tags = my_params.master_tags.HasTags();
// Options for read filtering & and trimming method selection
string trim_method = opts.GetFirstString ('-', "tag-trim-method", "sloppy-trim");
if (trim_method == "sloppy-trim")
my_params.tag_trim_method = kSloppyTrim;
else if (trim_method == "strict-trim")
my_params.tag_trim_method = kStrictTrim;
else {
cerr << "MolecularTagTrimmer Error: Unknown tag trimming option " << trim_method << endl;
exit(EXIT_FAILURE);
}
string filter_method = opts.GetFirstString ('-', "tag-filter-method", "need-all");
if (filter_method == "need-all")
my_params.tag_filter_method = kneed_all_tags;
else if (filter_method == "need-prefix")
my_params.tag_filter_method = kneed_only_prefix_tag;
else if (filter_method == "need-suffix")
my_params.tag_filter_method = kneed_only_suffix_tag;
else {
cerr << "MolecularTagTrimmer Error: Unknown tag filtering option " << filter_method << endl;
exit(EXIT_FAILURE);
}
return my_params;
}
bool BaseCallerContext::SetKeyAndFlowOrder(OptArgs& opts, const char * FlowOrder, const int NumFlows)
{
flow_order.SetFlowOrder( opts.GetFirstString ('-', "flow-order", FlowOrder),
opts.GetFirstInt ('f', "flowlimit", NumFlows));
if (flow_order.num_flows() > NumFlows)
flow_order.SetNumFlows(NumFlows);
assert(flow_order.is_ok());
string lib_key = opts.GetFirstString ('-', "lib-key", "TCAG"); //! @todo Get default key from wells
string tf_key = opts.GetFirstString ('-', "tf-key", "ATCG");
lib_key = opts.GetFirstString ('-', "librarykey", lib_key); // Backward compatible opts
tf_key = opts.GetFirstString ('-', "tfkey", tf_key);
keys.resize(2);
keys[0].Set(flow_order, lib_key, "lib");
keys[1].Set(flow_order, tf_key, "tf");
return true;
};
void RecalibrationModel::Initialize(OptArgs& opts, vector<string> &bam_comments, const string & run_id, const ion::ChipSubset & chip_subset)
{
string model_file_name = opts.GetFirstString ('-', "model-file", "");
int model_threshold = opts.GetFirstInt('-', "recal-model-hp-thres", 4);
bool save_hpmodel = opts.GetFirstBoolean('-', "save-hpmodel", true);
bool diagonal_state_prog = opts.GetFirstBoolean('-', "diagonal-state-prog", false);
if (diagonal_state_prog)
model_file_name.clear();
if (InitializeModel(model_file_name, model_threshold) and save_hpmodel)
SaveModelFileToBamComments(model_file_name, bam_comments, run_id, chip_subset.GetColOffset(), chip_subset.GetRowOffset());
}
void ExtendParameters::SetupFileIO(OptArgs &opts) {
// freeBayes slot
fasta = opts.GetFirstString('r', "reference", "");
if (fasta.empty()) {
cerr << "Fatal ERROR: Reference file not specified via -r" << endl;
exit(1);
}
ValidateAndCanonicalizePath(fasta);
// freeBayes slot
variantPriorsFile = opts.GetFirstString('c', "input-vcf", "");
if (variantPriorsFile.empty()) {
cerr << "INFO: No input VCF (Hotspot) file specified via -c,--input-vcf" << endl;
}
else
ValidateAndCanonicalizePath(variantPriorsFile);
sseMotifsFileName = opts.GetFirstString('e', "error-motifs", "");
sseMotifsProvided = true;
if (sseMotifsFileName.empty()) {
sseMotifsProvided = false;
cerr << "INFO: Systematic error motif file not specified via -e" << endl;
}
else
ValidateAndCanonicalizePath(sseMotifsFileName);
opts.GetOption(bams, "", 'b', "input-bam");
if (bams.empty()) {
cerr << "FATAL ERROR: BAM file not specified via -b" << endl;
exit(-1);
}
for (unsigned int i_bam = 0; i_bam < bams.size(); ++i_bam)
ValidateAndCanonicalizePath(bams[i_bam]);
outputDir = opts.GetFirstString('O', "output-dir", ".");
ValidateAndCanonicalizePath(outputDir);
outputFile = opts.GetFirstString('o', "output-vcf", "");
if (outputFile.empty()) {
cerr << "Fatal ERROR: Output VCF filename not specified via -o" << endl;
exit(1);
}
// Are those file names?
postprocessed_bam = opts.GetFirstString('-', "postprocessed-bam", "");
sampleName = opts.GetFirstString('g', "sample-name", "");
force_sample_name = opts.GetFirstString('-', "force-sample-name", "");
}
void PhaseEstimator::InitializeFromOptArgs(OptArgs& opts)
{
phasing_estimator_ = opts.GetFirstString ('-', "phasing-estimator", "spatial-refiner-2");
string arg_cf_ie_dr = opts.GetFirstString ('-', "libcf-ie-dr", "");
residual_threshold_ = opts.GetFirstDouble ('-', "phasing-residual-filter", 1.0);
max_phasing_levels_ = opts.GetFirstInt ('-', "max-phasing-levels", max_phasing_levels_default_);
use_pid_norm_ = opts.GetFirstString ('-', "keynormalizer", "keynorm-old") == "keynorm-new";
windowSize_ = opts.GetFirstInt ('-', "window-size", DPTreephaser::kWindowSizeDefault_);
if (!arg_cf_ie_dr.empty()) {
phasing_estimator_ = "override";
result_regions_x_ = 1;
result_regions_y_ = 1;
result_cf_.assign(1, 0.0);
result_ie_.assign(1, 0.0);
result_dr_.assign(1, 0.0);
if (3 != sscanf (arg_cf_ie_dr.c_str(), "%f,%f,%f", &result_cf_[0], &result_ie_[0], &result_dr_[0])) {
fprintf (stderr, "Option Error: libcf-ie-dr %s\n", arg_cf_ie_dr.c_str());
exit (EXIT_FAILURE);
}
return; // --libcf-ie-dr overrides other phasing-related options
}
}
void RecalibrationModel::Initialize(OptArgs& opts)
{
is_enabled_ = false;
string model_file_name = opts.GetFirstString ('-', "model-file", "");
if (model_file_name.empty() or model_file_name == "off") {
printf("RecalibrationModel: disabled\n\n");
return;
}
ifstream model_file;
model_file.open(model_file_name.c_str());
if (model_file.fail()) {
printf("RecalibrationModel: disabled (cannot open %s)\n\n", model_file_name.c_str());
model_file.close();
return;
}
recalModelHPThres = opts.GetFirstInt('-', "recal-model-hp-thres", 4);
string comment_line;
getline(model_file, comment_line); //skip the comment time
int flowStart, flowEnd, flowSpan, xMin, xMax, xSpan, yMin, yMax, ySpan, max_hp_calibrated;
model_file >> flowStart >> flowEnd >> flowSpan >> xMin >> xMax >> xSpan >> yMin >> yMax >> ySpan >> max_hp_calibrated;
stratification.SetupRegion(xMin, xMax, xSpan, yMin, yMax, ySpan);
//calculate number of partitions and initialize the stratifiedAs and stratifiedBs
SetupStratification(flowStart,flowEnd, flowSpan,xMin,xMax,xSpan,yMin,yMax,ySpan,max_hp_calibrated);
//TODO: parse model_file into stratifiedAs and stratifiedBs
while (model_file.good()) {
float paramA, paramB;
int refHP;
char flowBase;
model_file >> flowBase >> flowStart >> flowEnd >> xMin >> xMax >> yMin >> yMax >> refHP >> paramA >> paramB;
//populate it to stratifiedAs and startifiedBs
int nucInd = NuctoInt(flowBase);
//boundary check
int offsetRegion = stratification.OffsetRegion(xMin,yMin);
FillIndexes(offsetRegion,nucInd, refHP, flowStart, flowEnd, paramA, paramB);
}
model_file.close();
printf("Recalibration: enabled (using calibration file %s)\n\n", model_file_name.c_str());
is_enabled_ = true;
if (recalModelHPThres > MAX_HPXLEN) is_enabled_ = false;
}
void ExtendParameters::SetFreeBayesParameters(OptArgs &opts, Json::Value& fb_params) {
// FreeBayes parameters
// primarily used in candidate generation
targets = opts.GetFirstString('t', "target-file", "");
trim_ampliseq_primers = opts.GetFirstBoolean('-', "trim-ampliseq-primers", false);
if (targets.empty() and trim_ampliseq_primers) {
cerr << "ERROR: --trim-ampliseq-primers enabled but no --target-file provided" << endl;
exit(1);
}
allowIndels = RetrieveParameterBool (opts, fb_params, '-', "allow-indels", true);
allowSNPs = RetrieveParameterBool (opts, fb_params, '-', "allow-snps", true);
allowMNPs = RetrieveParameterBool (opts, fb_params, '-', "allow-mnps", true);
allowComplex = RetrieveParameterBool (opts, fb_params, '-', "allow-complex", false);
// deprecated:
// leftAlignIndels = RetrieveParameterBool (opts, fb_params, '-', "left-align-indels", false);
RetrieveParameterBool (opts, fb_params, '-', "left-align-indels", false);
//useBestNAlleles = 0;
useBestNAlleles = RetrieveParameterInt (opts, fb_params, 'm', "use-best-n-alleles", 2);
onlyUseInputAlleles = RetrieveParameterBool (opts, fb_params, '-', "use-input-allele-only", false);
min_mapping_qv = RetrieveParameterInt (opts, fb_params, 'M', "min-mapping-qv", 4);
read_snp_limit = RetrieveParameterInt (opts, fb_params, 'U', "read-snp-limit", 10);
readMaxMismatchFraction = RetrieveParameterDouble(opts, fb_params, 'z', "read-max-mismatch-fraction", 1.0);
maxComplexGap = RetrieveParameterInt (opts, fb_params, '!', "max-complex-gap", 1);
// read from json or command line, otherwise default to snp frequency
minAltFraction = RetrieveParameterDouble(opts, fb_params, '-', "gen-min-alt-allele-freq", my_controls.filter_snps.min_allele_freq);
minCoverage = RetrieveParameterInt (opts, fb_params, '-', "gen-min-coverage", my_controls.filter_snps.min_cov);
minIndelAltFraction = RetrieveParameterDouble(opts, fb_params, '-', "gen-min-indel-alt-allele-freq", my_controls.filter_hp_indel.min_allele_freq);
//set up debug levels
if (program_flow.DEBUG > 0)
debug = true;
if (program_flow.inputPositionsOnly) {
processInputPositionsOnly = true;
}
if (variantPriorsFile.empty() && (processInputPositionsOnly || onlyUseInputAlleles) ) {
cerr << "ERROR: Parameter error - Process-input-positions-only: " << processInputPositionsOnly << " use-input-allele-only: " << onlyUseInputAlleles << " : Specified without Input VCF File " << endl;
exit(1);
}
}
string RetrieveParameterString(OptArgs &opts, Json::Value& json, char short_name, const string& long_name_hyphens, const string& default_value)
{
string long_name_underscores = GetRidOfDomainAndHyphens(long_name_hyphens);
string value = default_value;
string source = "builtin default";
if (json.isMember(long_name_underscores)) {
value = json[long_name_underscores].asCString();
source = "parameters json file";
}
if (opts.HasOption(short_name, long_name_hyphens)) {
value = opts.GetFirstString(short_name, long_name_hyphens, value);
source = "command line option";
}
cout << setw(35) << long_name_hyphens << " = " << setw(10) << value << " (string, " << source << ")" << endl;
return value;
}
void ExtendParameters::ParametersFromJSON(OptArgs &opts, Json::Value &tvc_params, Json::Value &freebayes_params, Json::Value ¶ms_meta) {
string parameters_file = opts.GetFirstString('-', "parameters-file", "");
Json::Value parameters_json(Json::objectValue);
if (not parameters_file.empty()) {
ifstream in(parameters_file.c_str(), ifstream::in);
if (!in.good()) {
fprintf(stderr, "[tvc] FATAL ERROR: cannot open %s\n", parameters_file.c_str());
exit(-1);
}
in >> parameters_json;
in.close();
if (parameters_json.isMember("pluginconfig"))
parameters_json = parameters_json["pluginconfig"];
tvc_params = parameters_json.get("torrent_variant_caller", Json::objectValue);
freebayes_params = parameters_json.get("freebayes", Json::objectValue);
params_meta = parameters_json.get("meta", Json::objectValue);
}
int IonstatsReduceH5(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc-1, argv+1);
string output_h5_filename = opts.GetFirstString ('o', "output", "");
bool merge_proton_blocks = opts.GetFirstBoolean ('b', "merge-proton-blocks", "true");
vector<string> input_h5_filename;
opts.GetLeftoverArguments(input_h5_filename);
if(input_h5_filename.empty() or output_h5_filename.empty()) {
IonstatsReduceH5Help();
return 1;
}
if(merge_proton_blocks)
cout << "NOTE:" << argv[0] << " " << argv[1] << ": --merge-proton-blocks=true so any Proton block-specific read group suffixes will be merged" << endl;
return IonstatsAlignmentReduceH5(output_h5_filename, input_h5_filename, merge_proton_blocks);
}
bool BaseCallerParameters::InitializeFilesFromOptArgs(OptArgs& opts)
{
bc_files.input_directory = opts.GetFirstString ('i', "input-dir", ".");
bc_files.output_directory = opts.GetFirstString ('o', "output-dir", ".");
bc_files.unfiltered_untrimmed_directory = bc_files.output_directory + "/unfiltered.untrimmed";
bc_files.unfiltered_trimmed_directory = bc_files.output_directory + "/unfiltered.trimmed";
CreateResultsFolder ((char*)bc_files.output_directory.c_str());
CreateResultsFolder ((char*)bc_files.unfiltered_untrimmed_directory.c_str());
CreateResultsFolder ((char*)bc_files.unfiltered_trimmed_directory.c_str());
ValidateAndCanonicalizePath(bc_files.input_directory);
ValidateAndCanonicalizePath(bc_files.output_directory);
ValidateAndCanonicalizePath(bc_files.unfiltered_untrimmed_directory);
ValidateAndCanonicalizePath(bc_files.unfiltered_trimmed_directory);
bc_files.filename_wells = opts.GetFirstString ('-', "wells", bc_files.input_directory + "/1.wells");
bc_files.filename_mask = opts.GetFirstString ('-', "mask", bc_files.input_directory + "/analysis.bfmask.bin");
ValidateAndCanonicalizePath(bc_files.filename_wells);
ValidateAndCanonicalizePath(bc_files.filename_mask, bc_files.input_directory + "/bfmask.bin");
bc_files.filename_filter_mask = bc_files.output_directory + "/bfmask.bin";
bc_files.filename_json = bc_files.output_directory + "/BaseCaller.json";
bc_files.filename_phase = bc_files.output_directory + "/PhaseEstimates.json";
printf("\n");
printf("Input files summary:\n");
printf(" --input-dir %s\n", bc_files.input_directory.c_str());
printf(" --wells %s\n", bc_files.filename_wells.c_str());
printf(" --mask %s\n", bc_files.filename_mask.c_str());
printf("\n");
printf("Output directories summary:\n");
printf(" --output-dir %s\n", bc_files.output_directory.c_str());
printf(" unf.untr %s\n", bc_files.unfiltered_untrimmed_directory.c_str());
printf(" unf.tr %s\n", bc_files.unfiltered_trimmed_directory.c_str());
printf("\n");
bc_files.lib_datasets_file = opts.GetFirstString ('-', "datasets", "");
bc_files.calibration_panel_file = opts.GetFirstString ('-', "calibration-panel", "");
if (not bc_files.lib_datasets_file.empty())
ValidateAndCanonicalizePath(bc_files.lib_datasets_file);
if (not bc_files.calibration_panel_file.empty())
ValidateAndCanonicalizePath(bc_files.calibration_panel_file);
bc_files.options_set = true;
return true;
};
int IonstatsReduce(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc, argv);
string output_json_filename = opts.GetFirstString('o', "output", "");
vector<string> input_jsons;
opts.GetLeftoverArguments(input_jsons);
if(input_jsons.empty() or output_json_filename.empty()) {
IonstatsReduceHelp();
return 1;
}
ifstream in(input_jsons[0].c_str(), ifstream::in);
if (!in.good()) {
fprintf(stderr, "[ionstats] ERROR: cannot open %s\n", input_jsons[0].c_str());
return 1;
}
Json::Value first_input_json;
in >> first_input_json;
in.close();
if (!first_input_json.isMember("meta")) {
fprintf(stderr, "[ionstats] ERROR: %s is not a valid input file for ionstats reduce\n", input_jsons[0].c_str());
return 1;
}
string format_name = first_input_json["meta"].get("format_name","").asString();
if (format_name == "ionstats_basecaller")
return IonstatsBasecallerReduce(output_json_filename, input_jsons);
if (format_name == "ionstats_tf")
return IonstatsTestFragmentsReduce(output_json_filename, input_jsons);
if (format_name == "ionstats_alignment")
return IonstatsAlignmentReduce(output_json_filename, input_jsons);
fprintf(stderr, "[ionstats] ERROR: %s is not a valid input file for ionstats reduce\n", input_jsons[0].c_str());
return 1;
}
int main(int argc, const char* argv[])
{
printf ("tvcvalidator %s-%s (%s) - Prototype tvc validation tool\n\n",
IonVersion::GetVersion().c_str(), IonVersion::GetRelease().c_str(), IonVersion::GetSvnRev().c_str());
if (argc == 1) {
VariantValidatorHelp();
return 1;
}
OptArgs opts;
opts.ParseCmdLine(argc, argv);
if (opts.GetFirstBoolean('v', "version", false)) {
return 0;
}
if (opts.GetFirstBoolean('h', "help", false)) {
VariantValidatorHelp();
return 0;
}
string input_vcf_filename = opts.GetFirstString ('i', "input-vcf", "");
string truth_filename = opts.GetFirstString ('t', "truth-file", "");
string truth_dir = opts.GetFirstString ('d', "truth-dir", "/results/plugins/validateVariantCaller/files");
// TODO: reference optional, only used to verify reference allele in input-vcf and truth files
//string reference_filename = opts.GetFirstString ('r', "reference", "");
opts.CheckNoLeftovers();
//
// Step 1. Load input VCF file into memory
//
if (input_vcf_filename.empty()) {
VariantValidatorHelp();
cerr << "ERROR: Input VCF file not specified " << endl;
return 1;
}
VariantCallerResults results_vcf;
results_vcf.load_vcf(input_vcf_filename);
printf("Loaded VCF %s with %d variant calls\n", input_vcf_filename.c_str(), (int)results_vcf.variants.size());
//
// Step 2. Parse truth files, compare them to the input vcf, and compute match scores
//
if (not truth_filename.empty()) {
ValidatorTruth truth;
truth.ReadTruthFile(truth_filename);
truth.CompareToCalls(results_vcf);
return 0;
}
truth_dir += "/*.bed";
glob_t glob_result;
glob(truth_dir.c_str(), GLOB_TILDE, NULL, &glob_result);
for(unsigned int i = 0; i < glob_result.gl_pathc; ++i) {
ValidatorTruth truth;
truth.ReadTruthFile(string(glob_result.gl_pathv[i]));
truth.CompareToCalls(results_vcf);
}
globfree(&glob_result);
return 0;
}
int PrepareHotspots(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc, argv);
string input_bed_filename = opts.GetFirstString ('b', "input-bed", "");
string input_vcf_filename = opts.GetFirstString ('v', "input-vcf", "");
string input_real_vcf_filename = opts.GetFirstString ('p', "input-real-vcf", "");
string output_hot_vcf = opts.GetFirstString ('q', "output-fake-hot-vcf", "");
string output_bed_filename = opts.GetFirstString ('d', "output-bed", "");
string output_vcf_filename = opts.GetFirstString ('o', "output-vcf", "");
string reference_filename = opts.GetFirstString ('r', "reference", "");
string unmerged_bed = opts.GetFirstString ('u', "unmerged-bed", "");
bool left_alignment = opts.GetFirstBoolean('a', "left-alignment", false);
bool filter_bypass = opts.GetFirstBoolean('f', "filter-bypass", false);
bool allow_block_substitutions = opts.GetFirstBoolean('s', "allow-block-substitutions", true);
bool strict_check = opts.GetFirstBoolean('S', "strict-check", true);
opts.CheckNoLeftovers();
if((input_bed_filename.empty() == (input_vcf_filename.empty() and input_real_vcf_filename.empty())) or
(output_bed_filename.empty() and output_vcf_filename.empty()) or reference_filename.empty()) {
PrepareHotspotsHelp();
return 1;
}
if ((not input_real_vcf_filename.empty()) and (output_vcf_filename.empty() or not input_vcf_filename.empty())) {
PrepareHotspotsHelp();
return 1;
}
// Populate chromosome list from reference.fai
// Use mmap to fetch the entire reference
int ref_handle = open(reference_filename.c_str(),O_RDONLY);
struct stat ref_stat;
fstat(ref_handle, &ref_stat);
char *ref = (char *)mmap(0, ref_stat.st_size, PROT_READ, MAP_SHARED, ref_handle, 0);
FILE *fai = fopen((reference_filename+".fai").c_str(), "r");
if (!fai) {
fprintf(stderr, "ERROR: Cannot open %s.fai\n", reference_filename.c_str());
return 1;
}
vector<Reference> ref_index;
map<string,int> ref_map;
char line[1024], chrom_name[1024];
while (fgets(line, 1024, fai) != NULL) {
Reference ref_entry;
long chr_start;
if (5 != sscanf(line, "%1020s\t%ld\t%ld\t%d\t%d", chrom_name, &ref_entry.size, &chr_start,
&ref_entry.bases_per_line, &ref_entry.bytes_per_line))
continue;
ref_entry.chr = chrom_name;
ref_entry.start = ref + chr_start;
ref_index.push_back(ref_entry);
ref_map[ref_entry.chr] = (int) ref_index.size() - 1;
}
fclose(fai);
junction junc;
if (!unmerged_bed.empty()) {
FILE *fp = fopen(unmerged_bed.c_str(), "r");
if (!fp) {
fprintf(stderr, "ERROR: Cannot open %s\n", unmerged_bed.c_str());
return 1;
}
char line2[65536];
junc.init(ref_index.size());
bool line_overflow = false;
while (fgets(line2, 65536, fp) != NULL) {
if (line2[0] and line2[strlen(line2)-1] != '\n' and strlen(line2) == 65535) {
line_overflow = true;
continue;
}
if (line_overflow) {
line_overflow = false;
continue;
}
if (strstr(line2, "track")) continue;
char chr[100];
int b, e;
sscanf(line2, "%s %d %d", chr, &b, &e);
junc.add(ref_map[chr], b, e);
}
fclose(fp);
}
// Load input BED or load input VCF, group by chromosome
deque<LineStatus> line_status;
vector<deque<Allele> > alleles(ref_index.size());
if (!input_bed_filename.empty()) {
FILE *input = fopen(input_bed_filename.c_str(),"r");
if (!input) {
fprintf(stderr,"ERROR: Cannot open %s\n", input_bed_filename.c_str());
return 1;
}
char line2[65536];
int line_number = 0;
bool line_overflow = false;
while (fgets(line2, 65536, input) != NULL) {
if (line2[0] and line2[strlen(line2)-1] != '\n' and strlen(line2) == 65535) {
line_overflow = true;
continue;
}
line_number++;
if (line_overflow) {
line_overflow = false;
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot BED line: line length exceeds 64K";
continue;
}
if (strncmp(line2, "browser", 7) == 0)
continue;
if (strncmp(line2, "track", 5) == 0) {
if (string::npos != string(line2).find("allowBlockSubstitutions=true"))
allow_block_substitutions = true;
continue;
}
// OID= table has special meaning
if (string::npos != string(line2).find("OID=")) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Bed line contains OID=";
continue;
}
char *current_chr = strtok(line2, "\t\r\n");
char *current_start = strtok(NULL, "\t\r\n");
char *current_end = strtok(NULL, "\t\r\n");
char *current_id = strtok(NULL, "\t\r\n");
char *penultimate = strtok(NULL, "\t\r\n");
char *ultimate = strtok(NULL, "\t\r\n");
for (char *next = strtok(NULL, "\t\r\n"); next; next = strtok(NULL, "\t\r\n")) {
penultimate = ultimate;
ultimate = next;
}
if (!current_chr or !current_start or !current_end or !current_id or !penultimate or !ultimate) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot BED line: expected at least 6 fields";
continue;
}
Allele allele;
string string_chr(current_chr);
if (ref_map.find(string_chr) != ref_map.end())
allele.chr_idx = ref_map[string_chr];
else if (ref_map.find("chr"+string_chr) != ref_map.end())
allele.chr_idx = ref_map["chr"+string_chr];
else if (string_chr == "MT" and ref_map.find("chrM") != ref_map.end())
allele.chr_idx = ref_map["chrM"];
else {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Unknown chromosome name: ";
line_status.back().filter_message = string_chr;
continue;
}
allele.pos = strtol(current_start,NULL,10);
allele.id = current_id;
char *current_ref = NULL;
char *current_alt = NULL;
for (char *next = strtok(penultimate, ";"); next; next = strtok(NULL, ";")) {
if (strncmp(next,"REF=",4) == 0)
current_ref = next;
else if (strncmp(next,"OBS=",4) == 0)
current_alt = next;
else if (strncmp(next,"ANCHOR=",7) == 0) {
// ignore ANCHOR
} else {
char *value = next;
while (*value and *value != '=')
++value;
if (*value == '=')
*value++ = 0;
allele.custom_tags[next] = value;
}
}
if (!current_ref or !current_alt) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot BED line: REF and OBS fields required in penultimate column";
continue;
}
for (char *pos = current_ref+4; *pos; ++pos)
allele.ref += toupper(*pos);
for (char *pos = current_alt+4; *pos; ++pos)
allele.alt += toupper(*pos);
// here is the place to check the length of the hotspot cover the amplicon junction. ZZ
/*
if (junc.contain(allele.chr_idx, allele.pos, (unsigned int) allele.ref.size())) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "hotspot BED line contain the complete overlapping region of two amplicon, the variant cannot be detected by tvc";
continue;
}
if (not junc.contained_in_ampl(allele.chr_idx, allele.pos, (unsigned int) allele.ref.size())) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "hotspot BED line is not contained in any amplicon, the variant cannot be detected by tvc";
continue;
}
*/
allele.filtered = false;
line_status.push_back(LineStatus(line_number));
allele.line_status = &line_status.back();
allele.opos = allele.pos;
allele.oref = allele.ref;
allele.oalt = allele.alt;
alleles[allele.chr_idx].push_back(allele);
//line_status.back().allele = &alleles[allele.chr_idx].back();
line_status.back().chr_idx = allele.chr_idx;
line_status.back().opos = allele.opos;
line_status.back().id = allele.id;
}
fclose(input);
}
if (!input_vcf_filename.empty() or !input_real_vcf_filename.empty()) {
bool real_vcf = false;
FILE *input;
FILE *out_real = NULL;
FILE *out_hot = NULL;
int fake_ = 0;
int hn = 1;
if (!input_real_vcf_filename.empty()) {
real_vcf = true;
input = fopen(input_real_vcf_filename.c_str(),"r");
if (!input) {
fprintf(stderr,"ERROR: Cannot open %s\n", input_real_vcf_filename.c_str());
return 1;
}
out_real = fopen(output_vcf_filename.c_str(), "w");
if (!out_real) {
fprintf(stderr,"ERROR: Cannot open %s\n", output_vcf_filename.c_str());
return 1;
}
if (!output_hot_vcf.empty()) {
out_hot = fopen(output_hot_vcf.c_str(), "w");
if (!out_hot) {
fprintf(stderr,"ERROR: Cannot open %s\n", output_hot_vcf.c_str());
return 1;
}
} else out_hot = stdout;
fprintf(out_hot, "##fileformat=VCFv4.1\n##allowBlockSubstitutions=true\n#CHROM POS ID REF ALT QUAL FILTER INFO\n");
} else {
input = fopen(input_vcf_filename.c_str(),"r");
if (!input) {
fprintf(stderr,"ERROR: Cannot open %s\n", input_vcf_filename.c_str());
return 1;
}
}
char line2[65536];
char line3[65536];
int line_number = 0;
bool line_overflow = false;
list<one_vcfline> vcflist;
char last_chr[1024] = "";
while (fgets(line2, 65536, input) != NULL) {
if (line2[0] and line2[strlen(line2)-1] != '\n' and strlen(line2) == 65535) {
line_overflow = true;
continue;
}
line_number++;
if (line_overflow) {
line_overflow = false;
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot VCF line: line length exceeds 64K";
continue;
}
if (strncmp(line2, "##allowBlockSubstitutions=true", 30) == 0) {
allow_block_substitutions = true;
continue;
}
if (line2[0] == '#') {
if (out_real) { fprintf(out_real, "%s", line2);}
continue;
}
if (real_vcf) strcpy(line3, line2);
char *current_chr = strtok(line2, "\t\r\n");
char *current_start = strtok(NULL, "\t\r\n");
char *current_id = strtok(NULL, "\t\r\n");
char *current_ref = strtok(NULL, "\t\r\n");
char *current_alt = strtok(NULL, "\t\r\n");
strtok(NULL, "\t\r\n"); // Ignore QUAL
strtok(NULL, "\t\r\n"); // Ignore FILTER
char *current_info = strtok(NULL, "\t\r\n");
strtok(NULL, "\t\r\n");
char *gt = strtok(NULL, "\t\r\n");
if (!current_chr or !current_start or !current_id or !current_ref or !current_alt) {
line_status.push_back(LineStatus(line_number));
if (real_vcf) line_status.back().filter_message_prefix = "Malformed real VCF line: expected at least 5 fields";
else line_status.back().filter_message_prefix = "Malformed hotspot VCF line: expected at least 5 fields";
continue;
}
string string_chr(current_chr);
int chr_idx = 0;
if (ref_map.find(string_chr) != ref_map.end())
chr_idx = ref_map[string_chr];
else if (ref_map.find("chr"+string_chr) != ref_map.end())
chr_idx = ref_map["chr"+string_chr];
else if (string_chr == "MT" and ref_map.find("chrM") != ref_map.end())
chr_idx = ref_map["chrM"];
else {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Unknown chromosome name: ";
line_status.back().filter_message = string_chr;
continue;
}
for (char *pos = current_ref; *pos; ++pos)
*pos = toupper(*pos);
for (char *pos = current_alt; *pos; ++pos)
*pos = toupper(*pos);
// Process custom tags
vector<string> bstrand;
vector<string> hp_max_length;
string raw_oid;
string raw_omapalt;
string raw_oalt;
string raw_oref;
string raw_opos;
if (current_info) {
string raw_bstrand;
string raw_hp_max_length;
for (char *next = strtok(current_info, ";"); next; next = strtok(NULL, ";")) {
char *value = next;
while (*value and *value != '=')
++value;
if (*value == '=')
*value++ = 0;
if (strcmp(next, "TYPE") == 0)
continue;
if (strcmp(next, "HRUN") == 0)
continue;
if (strcmp(next, "HBASE") == 0)
continue;
if (strcmp(next, "FR") == 0)
continue;
if (strcmp(next, "OPOS") == 0) {
raw_opos = value;
continue;
}
if (strcmp(next, "OREF") == 0) {
raw_oref = value;
continue;
}
if (strcmp(next, "OALT") == 0) {
raw_oalt = value;
continue;
}
if (strcmp(next, "OID") == 0) {
raw_oid = value;
continue;
}
if (strcmp(next, "OMAPALT") == 0) {
raw_omapalt = value;
continue;
}
if (strcmp(next, "BSTRAND") == 0) {
raw_bstrand = value;
continue;
}
if (strcmp(next, "hp_max_length") == 0) {
raw_hp_max_length = value;
continue;
}
}
if (not raw_bstrand.empty())
split(raw_bstrand, ',', bstrand);
if (not raw_hp_max_length.empty())
split(raw_hp_max_length, ',', hp_max_length);
}
if (real_vcf) {
//fprintf(stderr, "%s\n", gt);
if (gt == NULL) continue;
// get gt
int g1 = atoi(gt), g2;
gt = strchr(gt, '/');
if (gt) g2 = atoi(gt+1);
else {fprintf(stderr, "GT not formatted right\n"); exit(1);}
//if (g1 == 0 and g2 == 0) continue;
unsigned int cur_pos = atoi(current_start);
one_vcfline newline(current_ref, current_alt, cur_pos, g1, g2, line3);
bool new_chr = false;
if (strcmp(current_chr, last_chr) != 0) {
new_chr = true;
}
while (not vcflist.empty()) {
if ((not new_chr) and vcflist.front().pos+strlen(vcflist.front().ref) > cur_pos) break;
if (vcflist.front().produce_hot_vcf(last_chr, out_real, hn, out_hot)) fake_++;
vcflist.pop_front();
}
if (new_chr) strcpy(last_chr, current_chr);
for (list<one_vcfline>::iterator it = vcflist.begin(); it != vcflist.end(); it++) {
it->check_subset(newline);
}
if (not newline.alts.empty()) vcflist.push_back(newline);
continue;
}
unsigned int allele_idx = 0;
for (char *sub_alt = strtok(current_alt,","); sub_alt; sub_alt = strtok(NULL,",")) {
Allele allele;
allele.chr_idx = chr_idx;
allele.ref = current_ref;
allele.alt = sub_alt;
allele.pos = strtol(current_start,NULL,10)-1;
allele.id = current_id;
if (allele.id == ".")
allele.id = "hotspot";
allele.filtered = false;
line_status.push_back(LineStatus(line_number));
allele.line_status = &line_status.back();
allele.opos = allele.pos;
allele.oref = allele.ref;
allele.oalt = allele.alt;
if (allele_idx < bstrand.size()) {
if (bstrand[allele_idx] != ".")
allele.custom_tags["BSTRAND"] = bstrand[allele_idx];
}
if (allele_idx < hp_max_length.size()) {
if (hp_max_length[allele_idx] != ".")
allele.custom_tags["hp_max_length"] = hp_max_length[allele_idx];
}
alleles[allele.chr_idx].push_back(allele);
//line_status.back().allele = &alleles[allele.chr_idx].back();
line_status.back().chr_idx = allele.chr_idx;
line_status.back().opos = allele.opos;
line_status.back().id = allele.id;
allele_idx++;
}
}
fclose(input);
if (real_vcf) {
while (not vcflist.empty()) {
if (vcflist.front().produce_hot_vcf(last_chr, out_real, hn, out_hot)) fake_++;
vcflist.pop_front();
}
fclose(out_real);
fclose(out_hot);
if (fake_ > 0)
return 0;
else return 1;
}
}
// Process by chromosome:
// - Verify reference allele
// - Left align
// - Sort
// - Filter for block substitutions, write
FILE *output_vcf = NULL;
if (!output_vcf_filename.empty()) {
output_vcf = fopen(output_vcf_filename.c_str(), "w");
if (!output_vcf) {
fprintf(stderr,"ERROR: Cannot open %s for writing\n", output_vcf_filename.c_str());
return 1;
}
fprintf(output_vcf, "##fileformat=VCFv4.1\n");
if (allow_block_substitutions)
fprintf(output_vcf, "##allowBlockSubstitutions=true\n");
fprintf(output_vcf, "#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\n");
}
FILE *output_bed = NULL;
if (!output_bed_filename.empty()) {
output_bed = fopen(output_bed_filename.c_str(), "w");
if (!output_bed) {
fprintf(stderr,"ERROR: Cannot open %s for writing\n", output_bed_filename.c_str());
if (output_vcf)
fclose(output_vcf);
return 1;
}
if (allow_block_substitutions)
fprintf(output_bed, "track name=\"hotspot\" type=bedDetail allowBlockSubstitutions=true\n");
else
fprintf(output_bed, "track name=\"hotspot\" type=bedDetail\n");
}
for (int chr_idx = 0; chr_idx < (int)ref_index.size(); ++chr_idx) {
for (deque<Allele>::iterator A = alleles[chr_idx].begin(); A != alleles[chr_idx].end(); ++A) {
// check bed file
if (junc.contain(A->chr_idx, A->pos, (unsigned int) A->ref.size())) {
A->filtered = true;
A->line_status->filter_message_prefix = "hotspot BED line contain the complete overlapping region of two amplicon, the variant cannot be detected by tvc";
continue;
}
if (not junc.contained_in_ampl(A->chr_idx, A->pos, (unsigned int) A->ref.size())) {
A->filtered = true;
A->line_status->filter_message_prefix = "hotspot BED line is not contained in any amplicon, the variant cannot be detected by tvc";
continue;
}
// Invalid characters
bool valid = true;
for (const char *c = A->ref.c_str(); *c ; ++c)
if (*c != 'A' and *c != 'C' and *c != 'G' and *c != 'T')
valid = false;
for (const char *c = A->alt.c_str(); *c ; ++c)
if (*c != 'A' and *c != 'C' and *c != 'G' and *c != 'T')
valid = false;
if (not valid) {
A->filtered = true;
A->line_status->filter_message_prefix = "REF and/or ALT contain characters other than ACGT: ";
A->line_status->filter_message = "REF = " + A->ref + " ALT = " + A->alt;
continue;
}
// Filter REF == ALT
if (A->ref == A->alt) {
A->filtered = true;
A->line_status->filter_message_prefix = "REF and ALT alleles equal";
continue;
}
// Confirm reference allele.
string ref_expected;
for (int idx = 0; idx < (int) A->ref.size(); ++idx)
ref_expected += ref_index[chr_idx].base(A->pos + idx);
if (A->ref != ref_expected) {
A->filtered = true;
A->line_status->filter_message_prefix = "Provided REF allele does not match reference: ";
A->line_status->filter_message = "Expected " + ref_expected + ", found " + A->ref;
continue;
}
// Trim
int ref_start = 0;
int ref_end = A->ref.size();
int alt_end = A->alt.size();
// Option 1: trim all trailing bases;
//while(ref_end and alt_end and A->ref[ref_end-1] == A->alt[alt_end-1]) {
// --ref_end;
// --alt_end;
//}
// Option 2: trim all leading basees;
//while (ref_start < ref_end and ref_start < alt_end and A->ref[ref_start] == A->alt[ref_start])
// ++ref_start;
// Option 3: trim anchor base if vcf
if (!input_vcf_filename.empty()) {
if (ref_end and alt_end and (ref_end == 1 or alt_end == 1) and A->ref[0] == A->alt[0])
ref_start = 1;
}
A->pos += ref_start;
A->ref = A->ref.substr(ref_start, ref_end-ref_start);
A->alt = A->alt.substr(ref_start, alt_end-ref_start);
ref_end -= ref_start;
alt_end -= ref_start;
// Left align
if (left_alignment && A->custom_tags.find("BSTRAND") == A->custom_tags.end()) { // black list variant not to be left aligned.
string trailing;
int can_do = 0, need_do = 0;
int ref_end_orig= ref_end, alt_end_orig = alt_end;
while(ref_end and alt_end and A->ref[ref_end-1] == A->alt[alt_end-1]) {
ref_end--; alt_end--;
}
if (ref_end == 0 || alt_end == 0) {
can_do = need_do = 1; // indel type, ZZ
} else {
int tmp_start = ref_start;
int ref_end_0 = ref_end, alt_end_0 = alt_end; // end after remove trailing match ZZ
while (tmp_start < ref_end and tmp_start < alt_end and A->ref[tmp_start] == A->alt[tmp_start])
++tmp_start;
if (tmp_start == ref_end || tmp_start == alt_end) {
can_do = 1; need_do = 0; // indel but indel is not at the left. ZZ
} else {
ref_end--; alt_end--;
while(ref_end and alt_end and A->ref[ref_end-1] == A->alt[alt_end-1]) {
ref_end--; alt_end--;
}
if (ref_end == 0 || alt_end == 0) {
// complex with 1 bp MM at right end
can_do = need_do = 1;
if (ref_end + alt_end == 0) need_do = 0; // SNP
} else {
int tmp_start0 = tmp_start; // start after removing leading matches
tmp_start++;
while (tmp_start < ref_end_orig and tmp_start < alt_end_orig and A->ref[tmp_start] == A->alt[tmp_start])
tmp_start++;
if (tmp_start >= ref_end_0 || tmp_start >= alt_end_0 || ref_end <= tmp_start0 || alt_end <= tmp_start0) {
// 1MM plus indel in middle, by definition cannot move the indel left enough to change A->pos
can_do = 1; need_do = 0;
} // else real complex
}
}
}
if (!can_do or !need_do) {
// do nothing
// if !can_do need add some more DP
ref_end = ref_end_orig;
alt_end = alt_end_orig;
} else {
// left align the indel part, here either ref_end = 0 or alt_end = 0
int opos = A->pos;
while (A->pos > 0) {
char nuc = ref_index[chr_idx].base(A->pos-1);
if (ref_end > 0 and A->ref[ref_end-1] != nuc)
break;
if (alt_end > 0 and A->alt[alt_end-1] != nuc)
break;
A->ref = string(1,nuc) + A->ref;
A->alt = string(1,nuc) + A->alt;
A->pos--;
}
if (ref_end != ref_end_orig) {
// trailing part is aligned, the whole ref and alt need to be kept. ZZ
ref_end = A->ref.size();
alt_end = A->alt.size();
}
if (junc.contain(chr_idx, A->pos, ref_end) or not junc.contained_in_ampl(chr_idx, A->pos, ref_end)) {
// after left align the hotspot contain an overlap region, revert to the original ZZ
if (opos != A->pos) {
A->ref.erase(0, opos-A->pos);
A->alt.erase(0, opos-A->pos);
A->pos = opos;
ref_end = ref_end_orig;
alt_end = alt_end_orig;
}
}
}
}
A->ref.resize(ref_end);
A->alt.resize(alt_end);
// Filter block substitutions: take 1
if (ref_end > 0 and alt_end > 0 and ref_end != alt_end and not allow_block_substitutions and not filter_bypass) {
A->filtered = true;
A->line_status->filter_message_prefix = "Block substitutions not supported";
continue;
}
}
if (output_bed) {
// Sort - without anchor base
stable_sort(alleles[chr_idx].begin(), alleles[chr_idx].end(), compare_alleles);
// Write
for (deque<Allele>::iterator I = alleles[chr_idx].begin(); I != alleles[chr_idx].end(); ++I) {
if (I->filtered)
continue;
fprintf(output_bed, "%s\t%ld\t%ld\t%s\tREF=%s;OBS=%s",
ref_index[chr_idx].chr.c_str(), I->pos, I->pos + I->ref.size(), I->id.c_str(),
I->ref.c_str(), I->alt.c_str());
for (map<string,string>::iterator C = I->custom_tags.begin(); C != I->custom_tags.end(); ++C)
fprintf(output_bed, ";%s=%s", C->first.c_str(), C->second.c_str());
fprintf(output_bed, "\tNONE\n");
/*
if (I->pos)
fprintf(output_bed, "%s\t%ld\t%ld\t%s\t0\t+\tREF=%s;OBS=%s;ANCHOR=%c\tNONE\n",
ref_index[chr_idx].chr.c_str(), I->pos, I->pos + I->ref.size(), I->id.c_str(),
I->ref.c_str(), I->alt.c_str(), ref_index[chr_idx].base(I->pos-1));
else
fprintf(output_bed, "%s\t%ld\t%ld\t%s\t0\t+\tREF=%s;OBS=%s;ANCHOR=\tNONE\n",
ref_index[chr_idx].chr.c_str(), I->pos, I->pos + I->ref.size(), I->id.c_str(),
I->ref.c_str(), I->alt.c_str());
*/
}
}
if (output_vcf) {
// Add anchor base to indels
for (deque<Allele>::iterator I = alleles[chr_idx].begin(); I != alleles[chr_idx].end(); ++I) {
if (I->filtered)
continue;
if (not I->ref.empty() and not I->alt.empty())
continue;
if (I->pos == 0) {
I->filtered = true;
I->line_status->filter_message_prefix = "INDELs at chromosome start not supported";
continue;
}
I->pos--;
I->ref = string(1,ref_index[chr_idx].base(I->pos)) + I->ref;
I->alt = string(1,ref_index[chr_idx].base(I->pos)) + I->alt;
}
// Sort - with anchor base
stable_sort(alleles[chr_idx].begin(), alleles[chr_idx].end(), compare_alleles);
// Merge alleles, remove block substitutions, write
for (deque<Allele>::iterator A = alleles[chr_idx].begin(); A != alleles[chr_idx].end(); ) {
string max_ref;
deque<Allele>::iterator B = A;
for (; B != alleles[chr_idx].end() and B->pos == A->pos; ++B)
if (!B->filtered and max_ref.size() < B->ref.size())
max_ref = B->ref;
bool filtered = true;
map<string,set<string> > unique_alts_and_ids;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
string new_alt = I->alt + max_ref.substr(I->ref.size());
if (new_alt.size() > 1 and max_ref.size() > 1 and new_alt.size() != max_ref.size() and not allow_block_substitutions and not filter_bypass) {
I->filtered = true;
I->line_status->filter_message_prefix = "Block substitutions not supported (post-merge)";
continue;
}
I->ref = max_ref;
I->alt = new_alt;
// Filter alleles with duplicate ALT + ID pairs
map<string,set<string> >::iterator alt_iter = unique_alts_and_ids.find(new_alt);
if (alt_iter != unique_alts_and_ids.end()) {
if (alt_iter->second.count(I->id) > 0) {
I->filtered = true;
I->line_status->filter_message_prefix = "Duplicate allele and ID";
continue;
}
}
unique_alts_and_ids[new_alt].insert(I->id);
filtered = false;
}
if (not filtered) {
fprintf(output_vcf, "%s\t%ld\t.\t%s\t",
ref_index[chr_idx].chr.c_str(), A->pos+1, max_ref.c_str());
bool comma = false;
map<string,map<string,string> > unique_alts_and_tags;
set<string> unique_tags;
set<string> unique_alt_alleles;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
unique_alts_and_tags[I->alt].insert(I->custom_tags.begin(), I->custom_tags.end());
for (map<string,string>::iterator S = I->custom_tags.begin(); S != I->custom_tags.end(); ++S)
unique_tags.insert(S->first);
if (unique_alt_alleles.count(I->alt) > 0)
continue;
unique_alt_alleles.insert(I->alt);
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->alt.c_str());
}
/*
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
map<string,map<string,string> >::iterator Q = unique_alts_and_tags.find(I->alt);
if (comma)
fprintf(output_vcf, ",");
comma = true;
if (Q == unique_alts_and_tags.end()) {fprintf(output_vcf, "."); continue;}
fprintf(output_vcf, "%s", Q->first.c_str());
}
*/
fprintf(output_vcf, "\t.\t.\tOID=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->id.c_str());
}
fprintf(output_vcf, ";OPOS=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%ld", I->opos+1);
}
fprintf(output_vcf, ";OREF=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->oref.c_str());
}
fprintf(output_vcf, ";OALT=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->oalt.c_str());
}
fprintf(output_vcf, ";OMAPALT=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->alt.c_str());
}
for (set<string>::iterator S = unique_tags.begin(); S != unique_tags.end(); ++S) {
fprintf(output_vcf, ";%s=", S->c_str());
comma=false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
map<string,map<string,string> >::iterator Q = unique_alts_and_tags.find(I->alt);
if (comma)
fprintf(output_vcf, ",");
comma = true;
if (Q == unique_alts_and_tags.end()) {fprintf(output_vcf, "."); continue;}
map<string,string>::iterator W = Q->second.find(*S);
if (W == Q->second.end())
fprintf(output_vcf, ".");
else
fprintf(output_vcf, "%s", W->second.c_str());
}
}
// fprintf(output_vcf, ";%s=%s", S->first.c_str(), S->second.c_str());
fprintf(output_vcf, "\n");
}
A = B;
}
}
}
if (output_bed) {
fflush(output_bed);
fclose(output_bed);
}
if (output_vcf) {
fflush(output_vcf);
fclose(output_vcf);
}
int lines_ignored = 0;
for (deque<LineStatus>::iterator L = line_status.begin(); L != line_status.end(); ++L) {
if (L->filter_message_prefix) {
if (L->chr_idx >= 0)
printf("Line %d ignored: [%s:%ld %s] %s%s\n", L->line_number, ref_index[L->chr_idx].chr.c_str(), L->opos+1, L->id.c_str(),
L->filter_message_prefix, L->filter_message.c_str());
else
printf("Line %d ignored: %s%s\n", L->line_number, L->filter_message_prefix, L->filter_message.c_str());
lines_ignored++;
}
}
printf("Ignored %d out of %d lines\n", lines_ignored, (int)line_status.size());
munmap(ref, ref_stat.st_size);
close(ref_handle);
if (lines_ignored > 0 and strict_check) return 1;
return 0;
}
int main (int argc, const char *argv[])
{
printf ("------------- bamrealignment --------------\n");
OptArgs opts;
opts.ParseCmdLine(argc, argv);
vector<int> score_vals(4);
string input_bam = opts.GetFirstString ('i', "input", "");
string output_bam = opts.GetFirstString ('o', "output", "");
opts.GetOption(score_vals, "4,-6,-5,-2", 's', "scores");
int clipping = opts.GetFirstInt ('c', "clipping", 2);
bool anchors = opts.GetFirstBoolean ('a', "anchors", true);
int bandwidth = opts.GetFirstInt ('b', "bandwidth", 10);
bool verbose = opts.GetFirstBoolean ('v', "verbose", false);
bool debug = opts.GetFirstBoolean ('d', "debug", false);
int format = opts.GetFirstInt ('f', "format", 1);
int num_threads = opts.GetFirstInt ('t', "threads", 8);
string log_fname = opts.GetFirstString ('l', "log", "");
if (input_bam.empty() or output_bam.empty())
return PrintHelp();
opts.CheckNoLeftovers();
std::ofstream logf;
if (log_fname.size ())
{
logf.open (log_fname.c_str ());
if (!logf.is_open ())
{
fprintf (stderr, "bamrealignment: Failed to open log file %s\n", log_fname.c_str());
return 1;
}
}
BamReader reader;
if (!reader.Open(input_bam)) {
fprintf(stderr, "bamrealignment: Failed to open input file %s\n", input_bam.c_str());
return 1;
}
SamHeader header = reader.GetHeader();
RefVector refs = reader.GetReferenceData();
BamWriter writer;
writer.SetNumThreads(num_threads);
if (format == 1)
writer.SetCompressionMode(BamWriter::Uncompressed);
else
writer.SetCompressionMode(BamWriter::Compressed);
if (!writer.Open(output_bam, header, refs)) {
fprintf(stderr, "bamrealignment: Failed to open output file %s\n", output_bam.c_str());
return 1;
}
// The meat starts here ------------------------------------
if (verbose)
cout << "Verbose option is activated, each alignment will print to screen." << endl
<< " After a read hit RETURN to continue to the next one," << endl
<< " or press q RETURN to quit the program," << endl
<< " or press s Return to silence verbose," << endl
<< " or press c RETURN to continue printing without further prompt." << endl << endl;
unsigned int readcounter = 0;
unsigned int mapped_readcounter = 0;
unsigned int realigned_readcounter = 0;
unsigned int modified_alignment_readcounter = 0;
unsigned int pos_update_readcounter = 0;
unsigned int failed_clip_realigned_readcount = 0;
unsigned int already_perfect_readcount = 0;
unsigned int bad_md_tag_readcount = 0;
unsigned int error_recreate_ref_readcount = 0;
unsigned int error_clip_anchor_readcount = 0;
unsigned int error_sw_readcount = 0;
unsigned int error_unclip_readcount = 0;
unsigned int start_position_shift;
int orig_position;
int new_position;
string md_tag, new_md_tag, input = "x";
vector<CigarOp> new_cigar_data;
vector<MDelement> new_md_data;
bool position_shift = false;
time_t start_time = time(NULL);
Realigner aligner;
aligner.verbose_ = verbose;
aligner.debug_ = debug;
if (!aligner.SetScores(score_vals))
cout << "bamrealignment: Four scores need to be provided: match, mismatch, gap open, gap extend score!" << endl;
aligner.SetAlignmentBandwidth(bandwidth);
BamAlignment alignment;
while(reader.GetNextAlignment(alignment)){
readcounter ++;
position_shift = false;
if ( (readcounter % 100000) == 0 )
cout << "Processed " << readcounter << " reads. Elapsed time: " << (time(NULL) - start_time) << endl;
if (alignment.IsMapped()) {
orig_position = alignment.Position;
mapped_readcounter++;
aligner.SetClipping(clipping, !alignment.IsReverseStrand());
if (aligner.verbose_) {
cout << endl;
if (alignment.IsReverseStrand())
cout << "The read is from the reverse strand." << endl;
else
cout << "The read is from the forward strand." << endl;
}
if (!alignment.GetTag("MD", md_tag)) {
if (aligner.verbose_)
cout << "Warning: Skipping read " << alignment.Name << ". It is mapped but missing MD tag." << endl;
if (logf.is_open ())
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "MISSMD" << '\n';
bad_md_tag_readcount++;
} else if (aligner.CreateRefFromQueryBases(alignment.QueryBases, alignment.CigarData, md_tag, anchors)) {
bool clipfail = false;
if (Realigner::CR_ERR_CLIP_ANCHOR == aligner.GetCreateRefError ())
{
clipfail = true;
failed_clip_realigned_readcount ++;
}
if (!aligner.computeSWalignment(new_cigar_data, new_md_data, start_position_shift)) {
if (aligner.verbose_)
cout << "Error in the alignment! Not updating read information." << endl;
if (logf.is_open ())
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "SWERR" << '\n';
error_sw_readcount++;
writer.SaveAlignment(alignment); // Write alignment unchanged
continue;
}
if (!aligner.addClippedBasesToTags(new_cigar_data, new_md_data, alignment.QueryBases.size())) {
if (aligner.verbose_)
cout << "Error when adding clipped anchors back to tags! Not updating read information." << endl;
if (logf.is_open ())
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "UNCLIPERR" << '\n';
writer.SaveAlignment(alignment); // Write alignment unchanged
error_unclip_readcount ++;
continue;
}
new_md_tag = aligner.GetMDstring(new_md_data);
realigned_readcounter++;
// adjust start position of read
if (!aligner.LeftAnchorClipped() and start_position_shift != 0) {
new_position = aligner.updateReadPosition(alignment.CigarData, (int)start_position_shift, alignment.Position);
if (new_position != alignment.Position) {
pos_update_readcounter++;
position_shift = true;
alignment.Position = new_position;
}
}
if (position_shift || alignment.CigarData.size () != new_cigar_data.size () || md_tag != new_md_tag)
{
if (logf.is_open ())
{
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "MOD";
if (position_shift)
logf << "-SHIFT";
if (clipfail)
logf << " NOCLIP";
logf << '\n';
}
modified_alignment_readcounter++;
}
else
{
if (logf.is_open ())
{
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "UNMOD";
if (clipfail)
logf << " NOCLIP";
logf << '\n';
}
}
if (aligner.verbose_){
cout << alignment.Name << endl;
cout << "------------------------------------------" << endl;
// Wait for input to continue or quit program
if (input.size() == 0)
input = 'x';
else if (input[0] != 'c' and input[0] != 'C')
getline(cin, input);
if (input.size()>0){
if (input[0] == 'q' or input[0] == 'Q')
return 1;
else if (input[0] == 's' or input[0] == 'S')
aligner.verbose_ = false;
}
}
// Finally update alignment information
alignment.CigarData = new_cigar_data;
alignment.EditTag("MD", "Z" , new_md_tag);
} // end of CreateRef else if
else {
switch (aligner.GetCreateRefError ())
{
case Realigner::CR_ERR_RECREATE_REF:
if (logf.is_open ())
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "RECRERR" << '\n';
error_recreate_ref_readcount++;
break;
case Realigner::CR_ERR_CLIP_ANCHOR:
if (logf.is_open ())
logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "CLIPERR" << '\n';
error_clip_anchor_readcount++;
break;
default:
// On a good run this writes way too many reads to the log file - don't want to create a too large txt file
// if (logf.is_open ())
//logf << alignment.Name << '\t' << alignment.IsReverseStrand() << '\t' << alignment.RefID << '\t' << setfill ('0') << setw (8) << orig_position << '\t' << "PERFECT" << '\n';
already_perfect_readcount++;
break;
}
if (aligner.verbose_) {
cout << alignment.Name << endl;
cout << "------------------------------------------" << endl;
// Wait for input to continue or quit program
if (input.size() == 0)
input = 'x';
else if (input[0] != 'c' and input[0] != 'C')
getline(cin, input);
if (input.size()>0){
if (input[0] == 'q' or input[0] == 'Q')
return 1;
else if (input[0] == 's' or input[0] == 'S')
aligner.verbose_ = false;
}
}
}
// --- Debug output for Rajesh ---
if (debug && aligner.invalid_cigar_in_input) {
aligner.verbose_ = true;
cout << "Invalid cigar string / md tag pair in read " << alignment.Name << endl;
// Rerun reference generation to display error
aligner.CreateRefFromQueryBases(alignment.QueryBases, alignment.CigarData, md_tag, anchors);
aligner.verbose_ = verbose;
aligner.invalid_cigar_in_input = false;
}
// --- --- ---
} // end of if isMapped
writer.SaveAlignment(alignment);
} // end while loop over reads
if (aligner.invalid_cigar_in_input)
cerr << "WARNING bamrealignment: There were invalid cigar string / md tag pairs in the input bam file." << endl;
// ----------------------------------------------------------------
// program end -- output summary information
cout << " File: " << input_bam << endl
<< " Total reads: " << readcounter << endl
<< " Mapped reads: " << mapped_readcounter << endl;
if (bad_md_tag_readcount)
cout << " Skipped: bad MD tags: " << bad_md_tag_readcount << endl;
if (error_recreate_ref_readcount)
cout << " Skipped: unable to recreate ref: " << error_recreate_ref_readcount << endl;
if (error_clip_anchor_readcount)
cout << " Skipped: error clipping anchor: " << error_clip_anchor_readcount << endl;
cout << " Skipped: already perfect: " << already_perfect_readcount << endl
<< " Total reads realigned: " << mapped_readcounter - already_perfect_readcount - bad_md_tag_readcount - error_recreate_ref_readcount - error_clip_anchor_readcount << endl;
if (failed_clip_realigned_readcount)
cout << " (including " << failed_clip_realigned_readcount << " that failed to clip)" << endl;
if (error_sw_readcount)
cout << " Failed to complete SW alignment: " << error_sw_readcount << endl;
if (error_unclip_readcount)
cout << " Failed to unclip anchor: " << error_unclip_readcount << endl;
cout << " Succesfully realigned: " << realigned_readcounter << endl
<< " Modified alignments: " << modified_alignment_readcounter << endl
<< " Shifted position: " << pos_update_readcounter << endl;
cout << "Processing time: " << (time(NULL)-start_time) << " seconds." << endl;
cout << "INFO: The output BAM file may be unsorted." << endl;
cout << "------------------------------------------" << endl;
return 0;
}
int IonstatsAlignment(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc, argv);
string input_bam_filename = opts.GetFirstString('i', "input", "");
string output_json_filename = opts.GetFirstString('o', "output", "ionstats_alignment.json");
int histogram_length = opts.GetFirstInt ('h', "histogram-length", 400);
if(argc < 2 or input_bam_filename.empty()) {
IonstatsAlignmentHelp();
return 1;
}
//
// Prepare for metric calculation
//
BamReader input_bam;
if (!input_bam.Open(input_bam_filename)) {
fprintf(stderr, "[ionstats] ERROR: cannot open %s\n", input_bam_filename.c_str());
return 1;
}
ReadLengthHistogram called_histogram;
ReadLengthHistogram aligned_histogram;
ReadLengthHistogram AQ7_histogram;
ReadLengthHistogram AQ10_histogram;
ReadLengthHistogram AQ17_histogram;
ReadLengthHistogram AQ20_histogram;
ReadLengthHistogram AQ47_histogram;
SimpleHistogram error_by_position;
called_histogram.Initialize(histogram_length);
aligned_histogram.Initialize(histogram_length);
AQ7_histogram.Initialize(histogram_length);
AQ10_histogram.Initialize(histogram_length);
AQ17_histogram.Initialize(histogram_length);
AQ20_histogram.Initialize(histogram_length);
AQ47_histogram.Initialize(histogram_length);
error_by_position.Initialize(histogram_length);
BamAlignment alignment;
vector<char> MD_op;
vector<int> MD_len;
MD_op.reserve(1024);
MD_len.reserve(1024);
string MD_tag;
//
// Main loop over mapped reads in the input BAM
//
while(input_bam.GetNextAlignment(alignment)) {
// Record read length
called_histogram.Add(alignment.Length);
if (!alignment.IsMapped() or !alignment.GetTag("MD",MD_tag))
continue;
//
// Step 1. Parse MD tag
//
MD_op.clear();
MD_len.clear();
for (const char *MD_ptr = MD_tag.c_str(); *MD_ptr;) {
int item_length = 0;
if (*MD_ptr >= '0' and *MD_ptr <= '9') { // Its a match
MD_op.push_back('M');
for (; *MD_ptr and *MD_ptr >= '0' and *MD_ptr <= '9'; ++MD_ptr)
item_length = 10*item_length + *MD_ptr - '0';
} else {
if (*MD_ptr == '^') { // Its a deletion
MD_ptr++;
MD_op.push_back('D');
} else // Its a substitution
MD_op.push_back('X');
for (; *MD_ptr and *MD_ptr >= 'A' and *MD_ptr <= 'Z'; ++MD_ptr)
item_length++;
}
MD_len.push_back(item_length);
}
//
// Step 2. Synchronously scan through Cigar and MD, doing error accounting
//
int MD_idx = alignment.IsReverseStrand() ? MD_op.size()-1 : 0;
int cigar_idx = alignment.IsReverseStrand() ? alignment.CigarData.size()-1 : 0;
int increment = alignment.IsReverseStrand() ? -1 : 1;
int AQ7_bases = 0;
int AQ10_bases = 0;
int AQ17_bases = 0;
int AQ20_bases = 0;
int AQ47_bases = 0;
int num_bases = 0;
int num_errors = 0;
while (cigar_idx < (int)alignment.CigarData.size() and MD_idx < (int) MD_op.size() and cigar_idx >= 0 and MD_idx >= 0) {
if (alignment.CigarData[cigar_idx].Length == 0) { // Try advancing cigar
cigar_idx += increment;
continue;
}
if (MD_len[MD_idx] == 0) { // Try advancing MD
MD_idx += increment;
continue;
}
// Match
if (alignment.CigarData[cigar_idx].Type == 'M' and MD_op[MD_idx] == 'M') {
int advance = min((int)alignment.CigarData[cigar_idx].Length, MD_len[MD_idx]);
num_bases += advance;
alignment.CigarData[cigar_idx].Length -= advance;
MD_len[MD_idx] -= advance;
// Insertion (read has a base, reference doesn't)
} else if (alignment.CigarData[cigar_idx].Type == 'I') {
int advance = alignment.CigarData[cigar_idx].Length;
for (int cnt = 0; cnt < advance; ++cnt) {
error_by_position.Add(num_bases);
num_bases++;
num_errors++;
}
alignment.CigarData[cigar_idx].Length -= advance;
// Deletion (reference has a base, read doesn't)
} else if (alignment.CigarData[cigar_idx].Type == 'D' and MD_op[MD_idx] == 'D') {
int advance = min((int)alignment.CigarData[cigar_idx].Length, MD_len[MD_idx]);
for (int cnt = 0; cnt < advance; ++cnt) {
error_by_position.Add(num_bases);
num_errors++;
}
alignment.CigarData[cigar_idx].Length -= advance;
MD_len[MD_idx] -= advance;
// Substitution
} else if (MD_op[MD_idx] == 'X') {
int advance = min((int)alignment.CigarData[cigar_idx].Length, MD_len[MD_idx]);
for (int cnt = 0; cnt < advance; ++cnt) {
error_by_position.Add(num_bases);
num_bases++;
num_errors++;
}
alignment.CigarData[cigar_idx].Length -= advance;
MD_len[MD_idx] -= advance;
} else {
printf("ionstats alignment: Unexpected OP combination: %s Cigar=%c, MD=%c !\n",
alignment.Name.c_str(), alignment.CigarData[cigar_idx].Type, MD_op[MD_idx]);
break;
}
if (num_errors*5 <= num_bases) AQ7_bases = num_bases;
if (num_errors*10 <= num_bases) AQ10_bases = num_bases;
if (num_errors*50 <= num_bases) AQ17_bases = num_bases;
if (num_errors*100 <= num_bases) AQ20_bases = num_bases;
if (num_errors == 0) AQ47_bases = num_bases;
}
//
// Step 3. Profit
//
if (num_bases >= 20) aligned_histogram.Add(num_bases);
if (AQ7_bases >= 20) AQ7_histogram.Add(AQ7_bases);
if (AQ10_bases >= 20) AQ10_histogram.Add(AQ10_bases);
if (AQ17_bases >= 20) AQ17_histogram.Add(AQ17_bases);
if (AQ20_bases >= 20) AQ20_histogram.Add(AQ20_bases);
if (AQ47_bases >= 20) AQ47_histogram.Add(AQ47_bases);
}
input_bam.Close();
//
// Processing complete, generate ionstats_alignment.json
//
Json::Value output_json(Json::objectValue);
output_json["meta"]["creation_date"] = get_time_iso_string(time(NULL));
output_json["meta"]["format_name"] = "ionstats_alignment";
output_json["meta"]["format_version"] = "1.0";
called_histogram.SaveToJson(output_json["full"]);
aligned_histogram.SaveToJson(output_json["aligned"]);
AQ7_histogram.SaveToJson(output_json["AQ7"]);
AQ10_histogram.SaveToJson(output_json["AQ10"]);
AQ17_histogram.SaveToJson(output_json["AQ17"]);
AQ20_histogram.SaveToJson(output_json["AQ20"]);
AQ47_histogram.SaveToJson(output_json["AQ47"]);
error_by_position.SaveToJson(output_json["error_by_position"]);
ofstream out(output_json_filename.c_str(), ios::out);
if (out.good()) {
out << output_json.toStyledString();
return 0;
} else {
fprintf(stderr, "ERROR: unable to write to '%s'\n", output_json_filename.c_str());
return 1;
}
return 0;
}
int IonstatsBasecaller(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc, argv);
string input_bam_filename = opts.GetFirstString('i', "input", "");
string output_json_filename = opts.GetFirstString('o', "output", "ionstats_basecaller.json");
int histogram_length = opts.GetFirstInt ('h', "histogram-length", 400);
if(argc < 2 or input_bam_filename.empty()) {
IonstatsBasecallerHelp();
return 1;
}
BamReader input_bam;
if (!input_bam.Open(input_bam_filename)) {
fprintf(stderr, "[ionstats] ERROR: cannot open %s\n", input_bam_filename.c_str());
return 1;
}
SamHeader sam_header = input_bam.GetHeader();
if(!sam_header.HasReadGroups()) {
fprintf(stderr, "[ionstats] ERROR: no read groups in %s\n", input_bam_filename.c_str());
return 1;
}
ReadLengthHistogram total_full_histo;
ReadLengthHistogram total_insert_histo;
ReadLengthHistogram total_Q17_histo;
ReadLengthHistogram total_Q20_histo;
total_full_histo.Initialize(histogram_length);
total_insert_histo.Initialize(histogram_length);
total_Q17_histo.Initialize(histogram_length);
total_Q20_histo.Initialize(histogram_length);
MetricGeneratorSNR system_snr;
BaseQVHistogram qv_histogram;
string flow_order;
string key;
for (SamReadGroupIterator rg = sam_header.ReadGroups.Begin(); rg != sam_header.ReadGroups.End(); ++rg) {
if(rg->HasFlowOrder())
flow_order = rg->FlowOrder;
if(rg->HasKeySequence())
key = rg->KeySequence;
}
double qv_to_error_rate[256];
for (int qv = 0; qv < 256; qv++)
qv_to_error_rate[qv] = pow(10.0,-0.1*(double)qv);
BamAlignment alignment;
string read_group;
vector<uint16_t> flow_signal_fz(flow_order.length());
vector<int16_t> flow_signal_zm(flow_order.length());
while(input_bam.GetNextAlignment(alignment)) {
// Record read length
unsigned int full_length = alignment.Length;
total_full_histo.Add(full_length);
// Record insert length
int insert_length = 0;
if (alignment.GetTag("ZA",insert_length))
total_insert_histo.Add(insert_length);
// Compute and record Q17 and Q20
int Q17_length = 0;
int Q20_length = 0;
double num_accumulated_errors = 0.0;
for(int pos = 0; pos < alignment.Length; ++pos) {
num_accumulated_errors += qv_to_error_rate[(int)alignment.Qualities[pos] - 33];
if (num_accumulated_errors / (pos + 1) <= 0.02)
Q17_length = pos + 1;
if (num_accumulated_errors / (pos + 1) <= 0.01)
Q20_length = pos + 1;
}
total_Q17_histo.Add(Q17_length);
total_Q20_histo.Add(Q20_length);
// Record data for system snr
if(alignment.GetTag("ZM", flow_signal_zm))
system_snr.Add(flow_signal_zm, key.c_str(), flow_order);
else if(alignment.GetTag("FZ", flow_signal_fz))
system_snr.Add(flow_signal_fz, key.c_str(), flow_order);
// Record qv histogram
qv_histogram.Add(alignment.Qualities);
}
input_bam.Close();
Json::Value output_json(Json::objectValue);
output_json["meta"]["creation_date"] = get_time_iso_string(time(NULL));
output_json["meta"]["format_name"] = "ionstats_basecaller";
output_json["meta"]["format_version"] = "1.0";
system_snr.SaveToJson(output_json);
qv_histogram.SaveToJson(output_json);
total_full_histo.SaveToJson(output_json["full"]);
total_insert_histo.SaveToJson(output_json["insert"]);
total_Q17_histo.SaveToJson(output_json["Q17"]);
total_Q20_histo.SaveToJson(output_json["Q20"]);
ofstream out(output_json_filename.c_str(), ios::out);
if (out.good()) {
out << output_json.toStyledString();
return 0;
} else {
fprintf(stderr, "ERROR: unable to write to '%s'\n", output_json_filename.c_str());
return 1;
}
}
BaseCallerFilters::BaseCallerFilters(OptArgs& opts,
const string& _flowOrder, int _numFlows, const vector<KeySequence>& _keys, Mask *_maskPtr)
{
flowOrder = _flowOrder;
keypassFilter = opts.GetFirstBoolean('k', "keypass-filter", true);
percentPositiveFlowsFilterTFs = opts.GetFirstBoolean('-', "clonal-filter-tf", false);
clonalFilterTraining = opts.GetFirstBoolean('-', "clonal-filter-train", false);
clonalFilterSolving = opts.GetFirstBoolean('-', "clonal-filter-solve", false);
minReadLength = opts.GetFirstInt ('-', "min-read-length", 8);
cafieResFilterCalling = opts.GetFirstBoolean('-', "cr-filter", false);
cafieResFilterTFs = opts.GetFirstBoolean('-', "cr-filter-tf", false);
generate_bead_summary_ = opts.GetFirstBoolean('-', "bead-summary", false);
// TODO: get this to work right. May require "unwound" flow order, so incompatible with current wells.FlowOrder()
//flt_control.cafieResMaxValueByFlowOrder[std::string ("TACG") ] = 0.06; // regular flow order
//flt_control.cafieResMaxValueByFlowOrder[std::string ("TACGTACGTCTGAGCATCGATCGATGTACAGC") ] = 0.08; // xdb flow order
cafieResMaxValue = opts.GetFirstDouble('-', "cr-filter-max-value", 0.08);
// SFFTrim options
trim_adapter = opts.GetFirstString('-', "trim-adapter", "ATCACCGACTGCCCATAGAGAGGCTGAGAC");
trim_adapter_cutoff = opts.GetFirstDouble('-', "trim-adapter-cutoff", 0.0);
trim_adapter_closest = opts.GetFirstBoolean('-', "trim-adapter-pick-closest", false);
trim_qual_wsize = opts.GetFirstInt('-', "trim-qual-window-size", 30);
trim_qual_cutoff = opts.GetFirstDouble('-', "trim-qual-cutoff", 100.0);
trim_min_read_len = opts.GetFirstInt('-', "trim-min-read-len", 8);
// Validate options
if (minReadLength < 1) {
fprintf (stderr, "Option Error: min-read-length must specify a positive value (%d invalid).\n", minReadLength);
exit (EXIT_FAILURE);
}
if (cafieResMaxValue <= 0) {
fprintf (stderr, "Option Error: cr-filter-max-value must specify a positive value (%lf invalid).\n", cafieResMaxValue);
exit (EXIT_FAILURE);
}
keys = _keys;
numClasses = keys.size();
assert(numClasses == 2);
classFilterPolyclonal.resize(numClasses);
classFilterPolyclonal[0] = clonalFilterSolving;
classFilterPolyclonal[1] = clonalFilterSolving && percentPositiveFlowsFilterTFs;
classFilterHighResidual.resize(numClasses);
classFilterHighResidual[0] = cafieResFilterCalling;
classFilterHighResidual[1] = cafieResFilterCalling && cafieResFilterTFs;
string filter_beverly_args = opts.GetFirstString('-', "beverly-filter", "0.03,0.03,8");
if (filter_beverly_args == "off") {
filter_beverly_enabled_ = false; // Nothing, really
printf("Beverly filter: disabled, use --beverly-filter=filter_ratio,trim_ratio,min_length\n");
} else {
int stat = sscanf (filter_beverly_args.c_str(), "%f,%f,%d",
&filter_beverly_filter_ratio_,
&filter_beverly_trim_ratio_,
&filter_beverly_min_read_length_);
if (stat != 3) {
fprintf (stderr, "Option Error: beverly-filter %s\n", filter_beverly_args.c_str());
fprintf (stderr, "Usage: --beverly-filter=filter_ratio,trim_ratio,min_length\n");
exit (EXIT_FAILURE);
}
filter_beverly_enabled_ = true;
printf("Beverly filter: enabled, use --beverly-filter=off to disable\n");
printf("Beverly filter: filter_ratio = %1.5f\n", filter_beverly_filter_ratio_);
printf("Beverly filter: trim_ratio = %1.5f\n", filter_beverly_trim_ratio_);
printf("Beverly filter: min_length = %d\n", filter_beverly_min_read_length_);
}
maskPtr = _maskPtr;
numFlows = _numFlows;
filterMask.assign(maskPtr->H()*maskPtr->W(), kUninitialized);
}
void PerBaseQual::Init(OptArgs& opts, const string& chip_type, const string &output_directory, bool recalib)
{
if(phred_table_)
{
delete [] phred_table_;
phred_table_ = 0;
}
string phred_table_file = opts.GetFirstString ('-', "phred-table-file", "");
save_predictors_ = opts.GetFirstBoolean('-', "save-predictors", false);
// Determine the correct phred table filename to use
bool binTable = true;
if (phred_table_file.empty()) {
ChipIdDecoder::SetGlobalChipId(chip_type.c_str());
ChipIdEnum chip_id = ChipIdDecoder::GetGlobalChipId();
switch(chip_id){
case ChipId314:
phred_table_file = "phredTable.txt_314.binary";
break;
case ChipId316:
phred_table_file = "phredTable.txt_316.binary";
break;
case ChipId316v2:
phred_table_file = "phredTable.txt_318.binary";
break;
case ChipId318:
phred_table_file = "phredTable.txt_318.binary";
break;
case ChipId900: // Proton chip
phred_table_file = "phredTable.txt_900.binary";
break;
default:
phred_table_file = "phredTable.txt_314.binary";
fprintf(stderr, "PerBaseQual: No default phred table for chip_type=%s, trying %s instead\n",
chip_type.c_str(), phred_table_file.c_str());
break;
}
if (recalib)
{
phred_table_file = phred_table_file.substr(0, phred_table_file.length() - 7);
phred_table_file += ".Recal.binary";
}
char* full_filename = GetIonConfigFile(phred_table_file.c_str());
if(!full_filename)
{
printf("WARNING: cannot find binary phred table file %s, try to use non-binary phred table\n", phred_table_file.c_str());
phred_table_file = phred_table_file.substr(0, phred_table_file.length() - 7); // get rid of .binary
binTable = false;
char* full_filename2 = GetIonConfigFile(phred_table_file.c_str());
if(!full_filename2)
ION_ABORT("ERROR: Can't find phred table file " + phred_table_file);
phred_table_file = full_filename2;
free(full_filename2);
}
else
{
phred_table_file = full_filename;
free(full_filename);
}
}
cout << endl << "PerBaseQual::Init... phred_table_file=" << phred_table_file << endl;
binTable = hasBinaryExtension(phred_table_file);
// Load the phred table
if(binTable)
{
cout << endl << "PerBaseQual::Init... load binary phred_table_file=" << phred_table_file << endl;
vector<size_t> vNumCuts(kNumPredictors, 0);
if(H5Fis_hdf5(phred_table_file.c_str()) > 0)
{
hid_t root = H5Fopen(phred_table_file.c_str(), H5F_ACC_RDONLY, H5P_DEFAULT);
if(root < 0)
{
ION_ABORT("ERROR: cannot open HDF5 file " + phred_table_file);
}
hid_t grpQvTable = H5Gopen(root, "/QvTable", H5P_DEFAULT);
if (grpQvTable < 0)
{
H5Fclose(root);
ION_ABORT("ERROR: fail to open HDF5 group QvTable");
}
if(H5Aexists(grpQvTable, "NumPredictors") <= 0)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: HDF5 attribute NumPredictors does not exist");
}
hid_t attrNumPreds = H5Aopen(grpQvTable, "NumPredictors", H5P_DEFAULT);
if (attrNumPreds < 0)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: fail to open HDF5 attribute NumPredictors");
}
unsigned int numPredictors = 0;
herr_t ret = H5Aread(attrNumPreds, H5T_NATIVE_UINT, &numPredictors);
H5Aclose(attrNumPreds);
if(ret < 0 || numPredictors != (unsigned int)kNumPredictors)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: HDF5 attribute NumPredictors is wrong");
}
char buf[100];
for(size_t i = 0; i < (size_t)kNumPredictors; ++i)
{
offsets_.push_back(1);
sprintf(buf, "ThresholdsOfPredictor%d", (int)i);
if(H5Aexists(grpQvTable, buf) <= 0)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: HDF5 attribute ThresholdsOfPredictor does not exist");
}
hid_t attrCuts = H5Aopen(grpQvTable, buf, H5P_DEFAULT);
if (attrCuts < 0)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: fail to open HDF5 attribute ThresholdsOfPredictor");
}
hsize_t size = H5Aget_storage_size(attrCuts);
size /= sizeof(float);
float* fcuts = new float[size];
ret = H5Aread(attrCuts, H5T_NATIVE_FLOAT, fcuts);
H5Aclose(attrCuts);
if(ret < 0)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: fail to read HDF5 attribute ThresholdsOfPredictor");
}
vector<float> vCuts(size);
copy(fcuts, fcuts + size, vCuts.begin());
phred_cuts_.push_back(vCuts);
delete [] fcuts;
fcuts = 0;
}
hid_t dsQvs = H5Dopen(grpQvTable, "Qvs", H5P_DEFAULT);
if (dsQvs < 0)
{
H5Gclose(grpQvTable);
H5Fclose(root);
ION_ABORT("ERROR: fail to open HDF5 dataset Qvs");
}
hsize_t tbSize = H5Dget_storage_size(dsQvs);
phred_table_ = new unsigned char[tbSize];
ret = H5Dread(dsQvs, H5T_NATIVE_UCHAR, H5S_ALL, H5S_ALL, H5P_DEFAULT, phred_table_);
H5Dclose(dsQvs);
H5Gclose(grpQvTable);
H5Fclose(root);
if (ret < 0)
{
delete [] phred_table_;
phred_table_ = 0;
ION_ABORT("ERROR: fail to read HDF5 dataset Qvs");
}
}
else
{
printf("WARNING: binary phred table file %s is not a HDF5 file, try binary file mode.\n", phred_table_file.c_str());
ifstream source;
source.open(phred_table_file.c_str(), ios::in|ios::binary|ios::ate);
if (!source.is_open())
ION_ABORT("ERROR: Cannot open file: " + phred_table_file);
long totalSize = source.tellg();
char* tbBlock = new char [totalSize];
source.seekg (0, ios::beg);
source.read (tbBlock, totalSize);
source.close();
long headerSize = 0;
char* ptr = tbBlock;
int numPredictors = ptr[0]; //kNumPredictors
if(numPredictors != kNumPredictors)
{
delete [] tbBlock;
tbBlock = 0;
ION_ABORT("ERROR: Wrong number of predictors load from " + phred_table_file);
}
ptr += 4;
headerSize += 4;
for(int i = 0; i < kNumPredictors; ++i)
{
vNumCuts[i] = ptr[0];
ptr += 4;
headerSize += 4;
offsets_.push_back(1);
}
long tbSize = 1;
for(int i = 0; i < kNumPredictors; ++i)
{
vector<float> vCuts;
tbSize *= vNumCuts[i];
for(size_t j = 0; j < vNumCuts[i]; ++j)
{
float tmp;
memcpy(&tmp, ptr, 4);
vCuts.push_back(tmp);
ptr += 4;
headerSize += 4;
}
phred_cuts_.push_back(vCuts);
}
if(tbSize != (totalSize - headerSize))
{
delete [] tbBlock;
tbBlock = 0;
ION_ABORT("ERROR: Wrong QV table size");
}
phred_table_ = new unsigned char[tbSize];
memcpy(phred_table_, ptr, tbSize * sizeof(unsigned char));
delete [] tbBlock;
tbBlock = 0;
}
for(size_t i = kNumPredictors - 2; i > 0; --i)
{
offsets_[i] *= phred_cuts_[i + 1].size();
offsets_[i - 1] = offsets_[i];
}
offsets_[0] *= phred_cuts_[1].size();
}
else
{
ifstream source;
source.open(phred_table_file.c_str());
if (!source.is_open())
ION_ABORT("ERROR: Cannot open file: " + phred_table_file);
while (!source.eof()) {
string line;
getline(source, line);
if (line.empty())
break;
if (line[0] == '#')
continue;
stringstream strs(line);
float temp;
for (int k = 0; k < kNumPredictors; ++k) {
strs >> temp;
phred_thresholds_[k].push_back(temp);
}
strs >> temp; //skip n-th entry
strs >> temp;
phred_quality_.push_back(temp);
}
source.close();
for (int k = 0; k < kNumPredictors; ++k)
phred_thresholds_max_[k] = *max_element(phred_thresholds_[k].begin(), phred_thresholds_[k].end());
}
// Prepare for predictor dump here
if (save_predictors_) {
string predictors_filename = output_directory + "/Predictors.txt";
cout << endl << "Saving PerBaseQual predictors to file " << predictors_filename << endl << endl;
predictor_dump_.open(predictors_filename.c_str());
if (!predictor_dump_.is_open())
ION_ABORT("ERROR: Cannot open file: " + predictors_filename);
}
}
int PrepareHotspots(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc, argv);
string input_bed_filename = opts.GetFirstString ('b', "input-bed", "");
string input_vcf_filename = opts.GetFirstString ('v', "input-vcf", "");
string output_bed_filename = opts.GetFirstString ('d', "output-bed", "");
string output_vcf_filename = opts.GetFirstString ('o', "output-vcf", "");
string reference_filename = opts.GetFirstString ('r', "reference", "");
bool left_alignment = opts.GetFirstBoolean('a', "left-alignment", false);
bool filter_bypass = opts.GetFirstBoolean('f', "filter-bypass", false);
bool allow_block_substitutions = opts.GetFirstBoolean('s', "allow-block-substitutions", false);
opts.CheckNoLeftovers();
if((input_bed_filename.empty() == input_vcf_filename.empty()) or
(output_bed_filename.empty() and output_vcf_filename.empty()) or reference_filename.empty()) {
PrepareHotspotsHelp();
return 1;
}
// Populate chromosome list from reference.fai
// Use mmap to fetch the entire reference
int ref_handle = open(reference_filename.c_str(),O_RDONLY);
struct stat ref_stat;
fstat(ref_handle, &ref_stat);
char *ref = (char *)mmap(0, ref_stat.st_size, PROT_READ, MAP_SHARED, ref_handle, 0);
FILE *fai = fopen((reference_filename+".fai").c_str(), "r");
if (!fai) {
fprintf(stderr, "ERROR: Cannot open %s.fai\n", reference_filename.c_str());
return 1;
}
vector<Reference> ref_index;
map<string,int> ref_map;
char line[1024], chrom_name[1024];
while (fgets(line, 1024, fai) != NULL) {
Reference ref_entry;
long chr_start;
if (5 != sscanf(line, "%s\t%ld\t%ld\t%d\t%d", chrom_name, &ref_entry.size, &chr_start,
&ref_entry.bases_per_line, &ref_entry.bytes_per_line))
continue;
ref_entry.chr = chrom_name;
ref_entry.start = ref + chr_start;
ref_index.push_back(ref_entry);
ref_map[ref_entry.chr] = (int) ref_index.size() - 1;
}
fclose(fai);
// Load input BED or load input VCF, group by chromosome
deque<LineStatus> line_status;
vector<deque<Allele> > alleles(ref_index.size());
if (!input_bed_filename.empty()) {
FILE *input = fopen(input_bed_filename.c_str(),"r");
if (!input) {
fprintf(stderr,"ERROR: Cannot open %s\n", input_bed_filename.c_str());
return 1;
}
char line2[65536];
int line_number = 0;
bool line_overflow = false;
while (fgets(line2, 65536, input) != NULL) {
if (line2[0] and line2[strlen(line2)-1] != '\n' and strlen(line2) == 65535) {
line_overflow = true;
continue;
}
line_number++;
if (line_overflow) {
line_overflow = false;
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot BED line: line length exceeds 64K";
continue;
}
if (strncmp(line2, "browser", 7) == 0)
continue;
if (strncmp(line2, "track", 5) == 0) {
if (string::npos != string(line2).find("allowBlockSubstitutions=true"))
allow_block_substitutions = true;
continue;
}
char *current_chr = strtok(line2, "\t\r\n");
char *current_start = strtok(NULL, "\t\r\n");
char *current_end = strtok(NULL, "\t\r\n");
char *current_id = strtok(NULL, "\t\r\n");
char *penultimate = strtok(NULL, "\t\r\n");
char *ultimate = strtok(NULL, "\t\r\n");
for (char *next = strtok(NULL, "\t\r\n"); next; next = strtok(NULL, "\t\r\n")) {
penultimate = ultimate;
ultimate = next;
}
if (!current_chr or !current_start or !current_end or !current_id or !penultimate or !ultimate) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot BED line: expected at least 6 fields";
continue;
}
Allele allele;
string string_chr(current_chr);
if (ref_map.find(string_chr) != ref_map.end())
allele.chr_idx = ref_map[string_chr];
else if (ref_map.find("chr"+string_chr) != ref_map.end())
allele.chr_idx = ref_map["chr"+string_chr];
else if (string_chr == "MT" and ref_map.find("chrM") != ref_map.end())
allele.chr_idx = ref_map["chrM"];
else {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Unknown chromosome name: ";
line_status.back().filter_message = string_chr;
continue;
}
allele.pos = strtol(current_start,NULL,10);
allele.id = current_id;
char *current_ref = NULL;
char *current_alt = NULL;
for (char *next = strtok(penultimate, ";"); next; next = strtok(NULL, ";")) {
if (strncmp(next,"REF=",4) == 0)
current_ref = next;
else if (strncmp(next,"OBS=",4) == 0)
current_alt = next;
}
if (!current_ref or !current_alt) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot BED line: REF and OBS fields required in penultimate column";
continue;
}
for (char *pos = current_ref+4; *pos; ++pos)
allele.ref += toupper(*pos);
for (char *pos = current_alt+4; *pos; ++pos)
allele.alt += toupper(*pos);
allele.filtered = false;
line_status.push_back(LineStatus(line_number));
allele.line_status = &line_status.back();
allele.opos = allele.pos;
allele.oref = allele.ref;
allele.oalt = allele.alt;
alleles[allele.chr_idx].push_back(allele);
line_status.back().allele = &alleles[allele.chr_idx].back();
}
fclose(input);
}
if (!input_vcf_filename.empty()) {
FILE *input = fopen(input_vcf_filename.c_str(),"r");
if (!input) {
fprintf(stderr,"ERROR: Cannot open %s\n", input_vcf_filename.c_str());
return 1;
}
char line2[65536];
int line_number = 0;
bool line_overflow = false;
while (fgets(line2, 65536, input) != NULL) {
if (line2[0] and line2[strlen(line2)-1] != '\n' and strlen(line2) == 65535) {
line_overflow = true;
continue;
}
line_number++;
if (line_overflow) {
line_overflow = false;
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot VCF line: line length exceeds 64K";
continue;
}
if (strncmp(line2, "##allowBlockSubstitutions=true", 30) == 0) {
allow_block_substitutions = true;
continue;
}
if (line2[0] == '#')
continue;
char *current_chr = strtok(line2, "\t\r\n");
char *current_start = strtok(NULL, "\t\r\n");
char *current_id = strtok(NULL, "\t\r\n");
char *current_ref = strtok(NULL, "\t\r\n");
char *current_alt = strtok(NULL, "\t\r\n");
if (!current_chr or !current_start or !current_id or !current_ref or !current_alt) {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Malformed hotspot VCF line: expected at least 5 fields";
continue;
}
string string_chr(current_chr);
int chr_idx = 0;
if (ref_map.find(string_chr) != ref_map.end())
chr_idx = ref_map[string_chr];
else if (ref_map.find("chr"+string_chr) != ref_map.end())
chr_idx = ref_map["chr"+string_chr];
else if (string_chr == "MT" and ref_map.find("chrM") != ref_map.end())
chr_idx = ref_map["chrM"];
else {
line_status.push_back(LineStatus(line_number));
line_status.back().filter_message_prefix = "Unknown chromosome name: ";
line_status.back().filter_message = string_chr;
continue;
}
for (char *pos = current_ref; *pos; ++pos)
*pos = toupper(*pos);
for (char *pos = current_alt; *pos; ++pos)
*pos = toupper(*pos);
for (char *sub_alt = strtok(current_alt,","); sub_alt; sub_alt = strtok(NULL,",")) {
Allele allele;
allele.chr_idx = chr_idx;
allele.ref = current_ref;
allele.alt = sub_alt;
allele.pos = strtol(current_start,NULL,10)-1;
allele.id = current_id;
if (allele.id == ".")
allele.id = "hotspot";
allele.filtered = false;
line_status.push_back(LineStatus(line_number));
allele.line_status = &line_status.back();
allele.opos = allele.pos;
allele.oref = allele.ref;
allele.oalt = allele.alt;
alleles[allele.chr_idx].push_back(allele);
line_status.back().allele = &alleles[allele.chr_idx].back();
}
}
fclose(input);
}
// Process by chromosome:
// - Verify reference allele
// - Left align
// - Sort
// - Filter for block substitutions, write
FILE *output_vcf = NULL;
if (!output_vcf_filename.empty()) {
output_vcf = fopen(output_vcf_filename.c_str(), "w");
if (!output_vcf) {
fprintf(stderr,"ERROR: Cannot open %s for writing\n", output_vcf_filename.c_str());
return 1;
}
fprintf(output_vcf, "##fileformat=VCFv4.1\n");
if (allow_block_substitutions)
fprintf(output_vcf, "##allowBlockSubstitutions=true\n");
fprintf(output_vcf, "#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\n");
}
FILE *output_bed = NULL;
if (!output_bed_filename.empty()) {
output_bed = fopen(output_bed_filename.c_str(), "w");
if (!output_bed) {
fprintf(stderr,"ERROR: Cannot open %s for writing\n", output_bed_filename.c_str());
if (output_vcf)
fclose(output_vcf);
return 1;
}
if (allow_block_substitutions)
fprintf(output_bed, "track name=\"hotspot\" type=bedDetail allowBlockSubstitutions=true\n");
else
fprintf(output_bed, "track name=\"hotspot\" type=bedDetail\n");
}
for (int chr_idx = 0; chr_idx < (int)ref_index.size(); ++chr_idx) {
for (deque<Allele>::iterator A = alleles[chr_idx].begin(); A != alleles[chr_idx].end(); ++A) {
// Invalid characters
bool valid = true;
for (const char *c = A->ref.c_str(); *c ; ++c)
if (*c != 'A' and *c != 'C' and *c != 'G' and *c != 'T')
valid = false;
for (const char *c = A->alt.c_str(); *c ; ++c)
if (*c != 'A' and *c != 'C' and *c != 'G' and *c != 'T')
valid = false;
if (not valid) {
A->filtered = true;
A->line_status->filter_message_prefix = "REF and/or ALT contain characters other than ACGT: ";
A->line_status->filter_message = "REF = " + A->ref + " ALT = " + A->alt;
continue;
}
// Filter REF == ALT
if (A->ref == A->alt) {
A->filtered = true;
A->line_status->filter_message_prefix = "REF and ALT alleles equal";
continue;
}
// Confirm reference allele.
string ref_expected;
for (int idx = 0; idx < (int) A->ref.size(); ++idx)
ref_expected += ref_index[chr_idx].base(A->pos + idx);
if (A->ref != ref_expected) {
A->filtered = true;
A->line_status->filter_message_prefix = "Provided REF allele does not match reference: ";
A->line_status->filter_message = "Expected " + ref_expected + ", found " + A->ref;
continue;
}
// Trim
int ref_start = 0;
int ref_end = A->ref.size();
int alt_end = A->alt.size();
// Option 1: trim all trailing bases
//while(ref_end and alt_end and A->ref[ref_end-1] == A->alt[alt_end-1]) {
// --ref_end;
// --alt_end;
//}
// Option 2: trim all leading basees
//while (ref_start < ref_end and ref_start < alt_end and A->ref[ref_start] == A->alt[ref_start])
// ++ref_start;
// Option 3: trim anchor base if vcf
if (!input_vcf_filename.empty()) {
if (ref_end and alt_end and (ref_end == 1 or alt_end == 1) and A->ref[0] == A->alt[0])
ref_start = 1;
}
A->pos += ref_start;
A->ref = A->ref.substr(ref_start, ref_end-ref_start);
A->alt = A->alt.substr(ref_start, alt_end-ref_start);
ref_end -= ref_start;
alt_end -= ref_start;
// Left align
if (left_alignment) {
while (A->pos > 0) {
char nuc = ref_index[chr_idx].base(A->pos-1);
if (ref_end > 0 and A->ref[ref_end-1] != nuc)
break;
if (alt_end > 0 and A->alt[alt_end-1] != nuc)
break;
A->ref = string(1,nuc) + A->ref;
A->alt = string(1,nuc) + A->alt;
A->pos--;
}
}
A->ref.resize(ref_end);
A->alt.resize(alt_end);
// Filter block substitutions: take 1
if (ref_end > 0 and alt_end > 0 and ref_end != alt_end and not allow_block_substitutions and not filter_bypass) {
A->filtered = true;
A->line_status->filter_message_prefix = "Block substitutions not supported";
continue;
}
}
if (output_bed) {
// Sort - without anchor base
sort(alleles[chr_idx].begin(), alleles[chr_idx].end(), compare_alleles);
// Write
for (deque<Allele>::iterator I = alleles[chr_idx].begin(); I != alleles[chr_idx].end(); ++I) {
if (I->filtered)
continue;
if (I->pos)
fprintf(output_bed, "%s\t%ld\t%ld\t%s\t0\t+\tREF=%s;OBS=%s;ANCHOR=%c\tNONE\n",
ref_index[chr_idx].chr.c_str(), I->pos, I->pos + I->ref.size(), I->id.c_str(),
I->ref.c_str(), I->alt.c_str(), ref_index[chr_idx].base(I->pos-1));
else
fprintf(output_bed, "%s\t%ld\t%ld\t%s\t0\t+\tREF=%s;OBS=%s;ANCHOR=\tNONE\n",
ref_index[chr_idx].chr.c_str(), I->pos, I->pos + I->ref.size(), I->id.c_str(),
I->ref.c_str(), I->alt.c_str());
}
}
if (output_vcf) {
// Add anchor base to indels
for (deque<Allele>::iterator I = alleles[chr_idx].begin(); I != alleles[chr_idx].end(); ++I) {
if (I->filtered)
continue;
if (not I->ref.empty() and not I->alt.empty())
continue;
if (I->pos == 0) {
I->filtered = true;
I->line_status->filter_message_prefix = "INDELs at chromosome start not supported";
continue;
}
I->pos--;
I->ref = string(1,ref_index[chr_idx].base(I->pos)) + I->ref;
I->alt = string(1,ref_index[chr_idx].base(I->pos)) + I->alt;
}
// Sort - with anchor base
sort(alleles[chr_idx].begin(), alleles[chr_idx].end(), compare_alleles);
// Merge alleles, remove block substitutions, write
for (deque<Allele>::iterator A = alleles[chr_idx].begin(); A != alleles[chr_idx].end(); ) {
string max_ref;
deque<Allele>::iterator B = A;
for (; B != alleles[chr_idx].end() and B->pos == A->pos; ++B)
if (!B->filtered and max_ref.size() < B->ref.size())
max_ref = B->ref;
bool filtered = true;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
string new_alt = I->alt + max_ref.substr(I->ref.size());
if (new_alt.size() > 1 and max_ref.size() > 1 and new_alt.size() != max_ref.size() and not allow_block_substitutions and not filter_bypass) {
I->filtered = true;
I->line_status->filter_message_prefix = "Block substitutions not supported (post-merge)";
continue;
}
I->ref = max_ref;
I->alt = new_alt;
filtered = false;
}
if (not filtered) {
fprintf(output_vcf, "%s\t%ld\t.\t%s\t",
ref_index[chr_idx].chr.c_str(), A->pos+1, max_ref.c_str());
bool comma = false;
set<string> unique_alt_alleles;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (unique_alt_alleles.count(I->alt) > 0)
continue;
unique_alt_alleles.insert(I->alt);
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->alt.c_str());
}
fprintf(output_vcf, "\t.\t.\tOID=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->id.c_str());
}
fprintf(output_vcf, ";OPOS=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%ld", I->opos+1);
}
fprintf(output_vcf, ";OREF=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->oref.c_str());
}
fprintf(output_vcf, ";OALT=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->oalt.c_str());
}
fprintf(output_vcf, ";OMAPALT=");
comma = false;
for (deque<Allele>::iterator I = A; I != B; ++I) {
if (I->filtered)
continue;
if (comma)
fprintf(output_vcf, ",");
comma = true;
fprintf(output_vcf, "%s", I->alt.c_str());
}
fprintf(output_vcf, "\n");
}
A = B;
}
}
}
if (output_bed) {
fflush(output_bed);
fclose(output_bed);
}
if (output_vcf) {
fflush(output_vcf);
fclose(output_vcf);
}
int lines_ignored = 0;
for (deque<LineStatus>::iterator L = line_status.begin(); L != line_status.end(); ++L) {
if (L->filter_message_prefix) {
if (L->allele)
printf("Line %d ignored: [%s:%ld %s] %s%s\n", L->line_number, ref_index[L->allele->chr_idx].chr.c_str(), L->allele->opos+1, L->allele->id.c_str(),
L->filter_message_prefix, L->filter_message.c_str());
else
printf("Line %d ignored: %s%s\n", L->line_number, L->filter_message_prefix, L->filter_message.c_str());
lines_ignored++;
}
}
printf("Ignored %d out of %d lines\n", lines_ignored, (int)line_status.size());
munmap(ref, ref_stat.st_size);
close(ref_handle);
return 0;
}
int IonstatsTestFragments(int argc, const char *argv[])
{
OptArgs opts;
opts.ParseCmdLine(argc, argv);
string input_bam_filename = opts.GetFirstString('i', "input", "");
string fasta_filename = opts.GetFirstString('r', "ref", "");
string output_json_filename = opts.GetFirstString('o', "output", "ionstats_tf.json");
int histogram_length = opts.GetFirstInt ('h', "histogram-length", 400);
if(argc < 2 or input_bam_filename.empty() or fasta_filename.empty()) {
IonstatsTestFragmentsHelp();
return 1;
}
//
// Prepare for metric calculation
//
map<string,string> tf_sequences;
PopulateReferenceSequences(tf_sequences, fasta_filename);
BamReader input_bam;
if (!input_bam.Open(input_bam_filename)) {
fprintf(stderr, "[ionstats] ERROR: cannot open %s\n", input_bam_filename.c_str());
return 1;
}
int num_tfs = input_bam.GetReferenceCount();
SamHeader sam_header = input_bam.GetHeader();
if(!sam_header.HasReadGroups()) {
fprintf(stderr, "[ionstats] ERROR: no read groups in %s\n", input_bam_filename.c_str());
return 1;
}
string flow_order;
string key;
for (SamReadGroupIterator rg = sam_header.ReadGroups.Begin(); rg != sam_header.ReadGroups.End(); ++rg) {
if(rg->HasFlowOrder())
flow_order = rg->FlowOrder;
if(rg->HasKeySequence())
key = rg->KeySequence;
}
// Need these metrics stratified by TF.
vector<ReadLengthHistogram> called_histogram(num_tfs);
vector<ReadLengthHistogram> aligned_histogram(num_tfs);
vector<ReadLengthHistogram> AQ10_histogram(num_tfs);
vector<ReadLengthHistogram> AQ17_histogram(num_tfs);
vector<SimpleHistogram> error_by_position(num_tfs);
vector<MetricGeneratorSNR> system_snr(num_tfs);
vector<MetricGeneratorHPAccuracy> hp_accuracy(num_tfs);
for (int tf = 0; tf < num_tfs; ++tf) {
called_histogram[tf].Initialize(histogram_length);
aligned_histogram[tf].Initialize(histogram_length);
AQ10_histogram[tf].Initialize(histogram_length);
AQ17_histogram[tf].Initialize(histogram_length);
error_by_position[tf].Initialize(histogram_length);
}
vector<uint16_t> flow_signal_fz(flow_order.length());
vector<int16_t> flow_signal_zm(flow_order.length());
const RefVector& refs = input_bam.GetReferenceData();
// Missing:
// - hp accuracy - tough, copy verbatim from TFMapper?
BamAlignment alignment;
vector<char> MD_op;
vector<int> MD_len;
MD_op.reserve(1024);
MD_len.reserve(1024);
string MD_tag;
//
// Main loop over mapped reads in the input BAM
//
while(input_bam.GetNextAlignment(alignment)) {
if (!alignment.IsMapped() or !alignment.GetTag("MD",MD_tag))
continue;
// The check below eliminates unexpected alignments
if (alignment.IsReverseStrand() or alignment.Position > 5)
continue;
int current_tf = alignment.RefID;
//
// Step 1. Parse MD tag
//
MD_op.clear();
MD_len.clear();
for (const char *MD_ptr = MD_tag.c_str(); *MD_ptr;) {
int item_length = 0;
if (*MD_ptr >= '0' and *MD_ptr <= '9') { // Its a match
MD_op.push_back('M');
for (; *MD_ptr and *MD_ptr >= '0' and *MD_ptr <= '9'; ++MD_ptr)
item_length = 10*item_length + *MD_ptr - '0';
} else {
if (*MD_ptr == '^') { // Its a deletion
MD_ptr++;
MD_op.push_back('D');
} else // Its a substitution
MD_op.push_back('X');
for (; *MD_ptr and *MD_ptr >= 'A' and *MD_ptr <= 'Z'; ++MD_ptr)
item_length++;
}
MD_len.push_back(item_length);
}
//
// Step 2. Synchronously scan through Cigar and MD, doing error accounting
//
int MD_idx = alignment.IsReverseStrand() ? MD_op.size()-1 : 0;
int cigar_idx = alignment.IsReverseStrand() ? alignment.CigarData.size()-1 : 0;
int increment = alignment.IsReverseStrand() ? -1 : 1;
int AQ10_bases = 0;
int AQ17_bases = 0;
int num_bases = 0;
int num_errors = 0;
while (cigar_idx < (int)alignment.CigarData.size() and MD_idx < (int) MD_op.size() and cigar_idx >= 0 and MD_idx >= 0) {
if (alignment.CigarData[cigar_idx].Length == 0) { // Try advancing cigar
cigar_idx += increment;
continue;
}
if (MD_len[MD_idx] == 0) { // Try advancing MD
MD_idx += increment;
continue;
}
// Match
if (alignment.CigarData[cigar_idx].Type == 'M' and MD_op[MD_idx] == 'M') {
int advance = min((int)alignment.CigarData[cigar_idx].Length, MD_len[MD_idx]);
num_bases += advance;
alignment.CigarData[cigar_idx].Length -= advance;
MD_len[MD_idx] -= advance;
// Insertion (read has a base, reference doesn't)
} else if (alignment.CigarData[cigar_idx].Type == 'I') {
int advance = alignment.CigarData[cigar_idx].Length;
for (int cnt = 0; cnt < advance; ++cnt) {
error_by_position[current_tf].Add(num_bases);
num_bases++;
num_errors++;
}
alignment.CigarData[cigar_idx].Length -= advance;
// Deletion (reference has a base, read doesn't)
} else if (alignment.CigarData[cigar_idx].Type == 'D' and MD_op[MD_idx] == 'D') {
int advance = min((int)alignment.CigarData[cigar_idx].Length, MD_len[MD_idx]);
for (int cnt = 0; cnt < advance; ++cnt) {
error_by_position[current_tf].Add(num_bases);
num_errors++;
}
alignment.CigarData[cigar_idx].Length -= advance;
MD_len[MD_idx] -= advance;
// Substitution
} else if (MD_op[MD_idx] == 'X') {
int advance = min((int)alignment.CigarData[cigar_idx].Length, MD_len[MD_idx]);
for (int cnt = 0; cnt < advance; ++cnt) {
error_by_position[current_tf].Add(num_bases);
num_bases++;
num_errors++;
}
alignment.CigarData[cigar_idx].Length -= advance;
MD_len[MD_idx] -= advance;
} else {
printf("ionstats tf: Unexpected OP combination: %s Cigar=%c, MD=%c !\n",
alignment.Name.c_str(), alignment.CigarData[cigar_idx].Type, MD_op[MD_idx]);
break;
}
if (num_errors*10 <= num_bases) AQ10_bases = num_bases;
if (num_errors*50 <= num_bases) AQ17_bases = num_bases;
}
//
// Step 3. Profit
//
called_histogram[current_tf].Add(alignment.Length);
aligned_histogram[current_tf].Add(num_bases);
AQ10_histogram[current_tf].Add(AQ10_bases);
AQ17_histogram[current_tf].Add(AQ17_bases);
if(alignment.GetTag("ZM", flow_signal_zm))
system_snr[current_tf].Add(flow_signal_zm, key.c_str(), flow_order);
else if(alignment.GetTag("FZ", flow_signal_fz))
system_snr[current_tf].Add(flow_signal_fz, key.c_str(), flow_order);
// HP accuracy - keeping it simple
if (!alignment.IsReverseStrand()) {
string genome = key + tf_sequences[refs[current_tf].RefName];
string calls = key + alignment.QueryBases;
const char *genome_ptr = genome.c_str();
const char *calls_ptr = calls.c_str();
for (int flow = 0; flow < (int)flow_order.length() and *genome_ptr and *calls_ptr; ++flow) {
int genome_hp = 0;
int calls_hp = 0;
while (*genome_ptr == flow_order[flow]) {
genome_hp++;
genome_ptr++;
}
while (*calls_ptr == flow_order[flow]) {
calls_hp++;
calls_ptr++;
}
hp_accuracy[current_tf].Add(genome_hp, calls_hp);
}
}
}
//
// Processing complete, generate ionstats_tf.json
//
Json::Value output_json(Json::objectValue);
output_json["meta"]["creation_date"] = get_time_iso_string(time(NULL));
output_json["meta"]["format_name"] = "ionstats_tf";
output_json["meta"]["format_version"] = "1.0";
output_json["results_by_tf"] = Json::objectValue;
for (int tf = 0; tf < num_tfs; ++tf) {
if (aligned_histogram[tf].num_reads() < 1000)
continue;
called_histogram[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]["full"]);
aligned_histogram[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]["aligned"]);
AQ10_histogram[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]["AQ10"]);
AQ17_histogram[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]["AQ17"]);
error_by_position[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]["error_by_position"]);
system_snr[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]);
hp_accuracy[tf].SaveToJson(output_json["results_by_tf"][refs[tf].RefName]);
output_json["results_by_tf"][refs[tf].RefName]["sequence"] = tf_sequences[refs[tf].RefName];
}
input_bam.Close();
ofstream out(output_json_filename.c_str(), ios::out);
if (out.good()) {
out << output_json.toStyledString();
return 0;
} else {
fprintf(stderr, "ERROR: unable to write to '%s'\n", output_json_filename.c_str());
return 1;
}
}
void PhaseEstimator::InitializeFromOptArgs(OptArgs& opts, const ion::ChipSubset & chip_subset, const string & key_norm_method)
{
// Parse command line options
phasing_estimator_ = opts.GetFirstString ('-', "phasing-estimator", "spatial-refiner-2");
vector<double> cf_ie_dr = opts.GetFirstDoubleVector('-', "libcf-ie-dr", "");
vector<double> init_cf_ie_dr = opts.GetFirstDoubleVector('-', "initcf-ie-dr", "");
residual_threshold_ = opts.GetFirstDouble ('-', "phasing-residual-filter", 1.0);
max_phasing_levels_ = opts.GetFirstInt ('-', "max-phasing-levels", max_phasing_levels_default_);
num_fullchip_iterations_= opts.GetFirstInt ('-', "phasing-fullchip-iterations", 3);
num_region_iterations_ = opts.GetFirstInt ('-', "phasing-region-iterations", 1);
num_reads_per_region_ = opts.GetFirstInt ('-', "phasing-num-reads", 5000);
min_reads_per_region_ = opts.GetFirstInt ('-', "phasing-min-reads", 1000);
phase_file_name_ = opts.GetFirstString ('-', "phase-estimation-file", "");
normalization_string_ = opts.GetFirstString ('-', "phase-normalization", "adaptive");
key_norm_method_ = key_norm_method;
// Static member variables
norm_during_param_eval_ = opts.GetFirstBoolean('-', "phase-norm-during-eval", false);
windowSize_ = opts.GetFirstInt ('-', "window-size", DPTreephaser::kWindowSizeDefault_);
phasing_start_flow_ = opts.GetFirstInt ('-', "phasing-start-flow", 70);
phasing_end_flow_ = opts.GetFirstInt ('-', "phasing-end-flow", 150);
inclusion_threshold_ = opts.GetFirstDouble ('-', "phasing-signal-cutoff", 1.4);
maxfrac_negative_flows_ = opts.GetFirstDouble ('-', "phasing-norm-threshold", 0.2);
// Initialize chip size - needed for loading phase parameters
chip_size_x_ = chip_subset.GetChipSizeX();
chip_size_y_ = chip_subset.GetChipSizeY();
region_size_x_ = chip_subset.GetRegionSizeX();
region_size_y_ = chip_subset.GetRegionSizeY();
num_regions_x_ = chip_subset.GetNumRegionsX();
num_regions_y_ = chip_subset.GetNumRegionsY();
num_regions_ = chip_subset.NumRegions();
// Loading existing phase estimates from a file takes precedence over all other options
if (not phase_file_name_.empty()) {
have_phase_estimates_ = LoadPhaseEstimationTrainSubset(phase_file_name_);
if (have_phase_estimates_) {
phasing_estimator_ = "override";
printf("Phase estimator settings:\n");
printf(" phase file name : %s\n", phase_file_name_.c_str());
printf(" phase estimation mode : %s\n\n", phasing_estimator_.c_str());
return;
} else
cout << "PhaseEstimator Error loading TrainSubset from file " << phase_file_name_ << endl;
}
// Set phase parameters if provided by command line
if (!cf_ie_dr.empty()) {
if (cf_ie_dr.size() != 3){
cerr << "BaseCaller Option Error: libcf-ie-dr needs to be a comma separated vector of 3 values." << endl;
exit (EXIT_FAILURE);
}
SetPhaseParameters(cf_ie_dr.at(0), cf_ie_dr.at(1), cf_ie_dr.at(2));
return; // --libcf-ie-dr overrides other phasing-related options
}
// Set starting values for estimation
if (!init_cf_ie_dr.empty()) {
if (init_cf_ie_dr.size() != 3){
cerr << "BaseCaller Option Error: initcf-ie-dr needs to be a comma separated vector of 3 values." << endl;
exit (EXIT_FAILURE);
}
init_cf_ = init_cf_ie_dr.at(0);
init_ie_ = init_cf_ie_dr.at(1);
init_dr_ = init_cf_ie_dr.at(2);
}
if (phasing_start_flow_ >= phasing_end_flow_ or phasing_start_flow_ < 0) {
cerr << "BaseCaller Option Error: phasing-start-flow " << phasing_start_flow_
<< "needs to be positive and smaller than phasing-end-flow " << phasing_end_flow_ << endl;
exit (EXIT_FAILURE);
}
if (normalization_string_ == "adaptive")
norm_method_ = 1;
else if (normalization_string_ == "pid")
norm_method_ = 2;
else if (normalization_string_ == "variable")
norm_method_ = 3;
else if (normalization_string_ == "off")
norm_method_ = 4;
else
norm_method_ = 0; // "gain" and anythign else is default
printf("Phase estimator settings:\n");
printf(" phase file name : %s\n", phase_file_name_.c_str());
printf(" phase estimation mode : %s\n", phasing_estimator_.c_str());
printf(" initial cf,ie,dr values: %f,%f,%f\n", init_cf_,init_ie_,init_dr_);
printf(" reads per region target: %d-%d\n", min_reads_per_region_, num_reads_per_region_);
printf(" normalization method : %s\n", normalization_string_.c_str());
printf(" variable norm threshold: %f\n", maxfrac_negative_flows_);
printf("\n");
}
