static int query_chroms(char *fname)
{
const char **seq;
int i, nseq, ftype = file_type(fname);
if ( ftype & IS_TXT || !ftype )
{
tbx_t *tbx = tbx_index_load(fname);
if ( !tbx ) error("Could not load .tbi index of %s\n", fname);
seq = tbx_seqnames(tbx, &nseq);
for (i=0; i<nseq; i++)
printf("%s\n", seq[i]);
free(seq);
tbx_destroy(tbx);
}
else if ( ftype==IS_BCF )
{
htsFile *fp = hts_open(fname,"r");
if ( !fp ) error("Could not read %s\n", fname);
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) error("Could not read the header: %s\n", fname);
hts_close(fp);
hts_idx_t *idx = bcf_index_load(fname);
if ( !idx ) error("Could not load .csi index of %s\n", fname);
seq = bcf_index_seqnames(idx, hdr, &nseq);
for (i=0; i<nseq; i++)
printf("%s\n", seq[i]);
free(seq);
bcf_hdr_destroy(hdr);
hts_idx_destroy(idx);
}
else if ( ftype==IS_BAM ) // todo: BAM
error("BAM: todo\n");
return 0;
}
TBXOrderedReader::TBXOrderedReader(std::string hts_file, std::vector<GenomeInterval>& intervals)
{
this->hts_file = hts_file;
this->intervals = intervals;
interval_index = 0;
hts = NULL;
tbx = NULL;
itr = NULL;
s = {0, 0, 0};
hts = hts_open(hts_file.c_str(), "r");
intervals_present = intervals.size()!=0;
if ((tbx = tbx_index_load(hts_file.c_str())))
{
index_loaded = true;
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] index cannot be loaded for %s\n", __FUNCTION__, hts_file.c_str());
exit(1);
}
}
random_access_enabled = intervals_present && index_loaded;
};
int bcf_sr_set_targets(readers_t *files, const char *fname)
{
regions_t *tgts = (regions_t *) calloc(1,sizeof(regions_t));
tgts->file = hts_open(fname, "rb", NULL);
if ( !tgts->file ) return 0;
tgts->tbx = tbx_index_load(fname);
tgts->seq_names = (char**) tbx_seqnames(tgts->tbx, &tgts->nseqs);
tgts->cseq = -1;
files->targets = tgts;
return 1;
}
Tabix::Tabix(string& file) {
has_jumped = false;
filename = file;
const char* cfilename = file.c_str();
struct stat stat_tbi,stat_vcf;
char *fnidx = (char*) calloc(strlen(cfilename) + 5, 1);
strcat(strcpy(fnidx, cfilename), ".tbi");
if ( bgzf_is_bgzf(cfilename)!=1 )
{
cerr << "[tabix++] was bgzip used to compress this file? " << file << endl;
free(fnidx);
exit(1);
}
// Common source of errors: new VCF is used with an old index
stat(fnidx, &stat_tbi);
stat(cfilename, &stat_vcf);
if ( stat_vcf.st_mtime > stat_tbi.st_mtime )
{
cerr << "[tabix++] the index file is older than the vcf file. Please use '-f' to overwrite or reindex." << endl;
free(fnidx);
exit(1);
}
free(fnidx);
if ((fn = hts_open(cfilename, "r")) == 0) {
cerr << "[tabix++] fail to open the data file." << endl;
exit(1);
}
if ((tbx = tbx_index_load(cfilename)) == NULL) {
cerr << "[tabix++] failed to load the index file." << endl;
exit(1);
}
int nseq;
const char** seq = tbx_seqnames(tbx, &nseq);
for (int i=0; i<nseq; i++) {
chroms.push_back(seq[i]);
}
free(seq);
idxconf = &tbx_conf_vcf;
// set up the iterator, defaults to the beginning
if (nseq == 0){
// the vcf file contains only the header according
// to the index
iter = NULL;
current_chrom = chroms.end();
}else{
current_chrom = chroms.begin();
iter = tbx_itr_querys(tbx, current_chrom->c_str());
}
}
void
hts_streamer::
_load_index()
{
if (_tidx) return;
_tidx = tbx_index_load(name());
if (! _tidx)
{
std::ostringstream oss;
oss << "Failed to load index for hts file: '" << name() << "'";
BOOST_THROW_EXCEPTION(illumina::common::GeneralException(oss.str()));
}
}
void cis_data::scanPhenotypes(string fbed) {
int n_includedP = 0;
int n_excludedP = 0;
int n_negativeStrd = 0;
//Open BED file
vrb.title("Scanning phenotype data in [" + fbed + "]");
htsFile *fp = hts_open(fbed.c_str(),"r");
if (!fp) vrb.error("Cannot open file");
tbx_t * tbx = tbx_index_load(fbed.c_str());
if (!tbx) vrb.error("Cannot open index file");
//Read header
kstring_t str = {0,0,0};
if (!hts_getline(fp, KS_SEP_LINE, &str) || !str.l || str.s[0] != tbx->conf.meta_char ) vrb.error("Cannot read header line");
//Scan file
vector < string > tokens;
while (hts_getline(fp, KS_SEP_LINE, &str) >= 0) {
if (str.l && str.s[0] != tbx->conf.meta_char) {
stb.split(string(str.s), tokens);
if (tokens.size() < 5) vrb.error("Incorrect number of columns!");
if ((grp_mode == GRP_NONE && filter_phenotype.check(tokens[3])) || (grp_mode != GRP_NONE && filter_phenotype.check(tokens[4]))) {
phenotype_id.push_back(tokens[3]);
phenotype_chr.push_back(tokens[0]);
phenotype_start.push_back(atoi(tokens[1].c_str()) + 1);
phenotype_end.push_back(atoi(tokens[2].c_str()));
if (grp_mode > 0 && full_test) phenotype_grp.push_back("ALL_GENES");
if (grp_mode > 0 && !full_test) phenotype_grp.push_back(tokens[4]);
phenotype_neg.push_back(tokens[5] == "-");
if (phenotype_neg.back()) n_negativeStrd ++;
n_includedP++;
} else n_excludedP ++;
}
}
//Finalize & verbose
tbx_destroy(tbx);
if (hts_close(fp)) vrb.error("Cannot properly close file");
phenotype_count = phenotype_id.size();
vrb.bullet(stb.str(n_includedP) + " phenotypes included");
if (n_excludedP > 0) vrb.bullet(stb.str(n_excludedP) + " phenotypes excluded by user");
if (n_negativeStrd > 0 ) vrb.bullet(stb.str(n_negativeStrd) + " phenotypes are on the negative strand");
if (phenotype_count == 0) vrb.leave("Cannot find phenotypes in region!");
}
/**
* Load index for the ith file, returns true if successful
*/
bool BCFSyncedReader::load_index(int32_t i)
{
if (ftypes[i].format==bcf && ftypes[i].compression==bgzf)
{
if (!(idxs[i] = bcf_index_load(file_names[i].c_str())))
{
return false;
}
}
else if (ftypes[i].format==vcf && ftypes[i].compression==bgzf)
{
if (!(tbxs[i] = tbx_index_load(file_names[i].c_str())))
{
return false;
}
}
return true;
}
/**
* Load index for the ith file, returns true if successful
*/
bool BCFSyncedStreamReader::load_index(int32_t i)
{
if (ftypes[i]==FT_BCF_GZ)
{
if (!(idxs[i] = bcf_index_load(vcf_files[i].c_str())))
{
return false;
}
}
else if (ftypes[i]==FT_VCF_GZ)
{
if (!(tbxs[i] = tbx_index_load(vcf_files[i].c_str())))
{
return false;
}
}
return true;
}
int main(int argc, char* argv[]) {
namespace po = boost::program_options;
std::string file;
std::string output;
try
{
// Declare the supported options.
po::options_description desc("Allowed options");
desc.add_options()
("help,h", "produce help message")
("version", "Show version")
("input-file", po::value< std::string >(), "The input files")
("output-file", po::value<std::string>(), "The output file name.")
;
po::positional_options_description popts;
popts.add("input-file", 1);
popts.add("output-file", 1);
po::options_description cmdline_options;
cmdline_options
.add(desc)
;
po::variables_map vm;
po::store(po::command_line_parser(argc, argv).
options(cmdline_options).positional(popts).run(), vm);
po::notify(vm);
if (vm.count("version"))
{
std::cout << "vcfhdr2json version " << HAPLOTYPES_VERSION << "\n";
return 0;
}
if (vm.count("help"))
{
std::cout << desc << "\n";
return 1;
}
if (vm.count("input-file"))
{
file = vm["input-file"].as< std::string > ();
}
if (vm.count("output-file"))
{
output = vm["output-file"].as< std::string >();
}
if(file.size() == 0)
{
std::cerr << "Please specify an input file.\n";
return 1;
}
if (output == "")
{
std::cerr << "Please specify an output file.\n";
return 1;
}
}
catch (po::error & e)
{
std::cerr << e.what() << "\n";
return 1;
}
try
{
Json::StyledWriter writer;
htsFile * fp = bcf_open(file.c_str(), "r");
bcf_hdr_t * hdr = bcf_hdr_read(fp);
Json::Value root;
Json::Value a;
for (int i = 0; i < bcf_hdr_nsamples(hdr); ++i)
{
a.append(hdr->samples[i]);
}
root["samples"] = a;
Json::Value fields;
for (int i = 0; i < hdr->nhrec; i++)
{
Json::Value field;
field["key"] = hdr->hrec[i]->key;
if (!hdr->hrec[i]->value)
{
Json::Value values;
for (int j = 0; j < hdr->hrec[i]->nkeys; j++)
{
values[hdr->hrec[i]->keys[j]] = hdr->hrec[i]->vals[j];
}
field["values"] = values;
}
else
{
field["value"] = hdr->hrec[i]->value;
}
fields.append(field);
}
root["fields"] = fields;
tbx_t * tbx_idx = tbx_index_load(file.c_str());
if ( !tbx_idx )
{
hts_idx_t * csi_idx = bcf_index_load(file.c_str());
if(!csi_idx)
{
root["tabix"] = Json::Value::null;
}
else
{
root["tabix"] = Json::Value();
root["tabix"]["chromosomes"] = Json::Value();
int count = 0;
const char ** tbx_names = bcf_index_seqnames(csi_idx, hdr, &count);
for (int i = 0; i < count; ++i)
{
root["tabix"]["chromosomes"].append(tbx_names[i]);
}
free(tbx_names);
hts_idx_destroy(csi_idx);
}
}
else
{
root["tabix"] = Json::Value();
root["tabix"]["chromosomes"] = Json::Value();
int count = 0;
const char ** tbx_names = tbx_seqnames(tbx_idx, &count);
for (int i = 0; i < count; ++i)
{
root["tabix"]["chromosomes"].append(tbx_names[i]);
}
free(tbx_names);
tbx_destroy(tbx_idx);
}
std::ofstream out(output.c_str());
out << writer.write(root);
bcf_close(fp);
bcf_hdr_destroy(hdr);
}
catch(std::runtime_error & e)
{
std::cerr << e.what() << std::endl;
return 1;
}
catch(std::logic_error & e)
{
std::cerr << e.what() << std::endl;
return 1;
}
return 0;
}
bcf_hdr_t *vcf_hdr_read(htsFile *fp)
{
if (!fp->is_bin) {
kstring_t txt, *s = &fp->line;
bcf_hdr_t *h;
h = bcf_hdr_init();
txt.l = txt.m = 0; txt.s = 0;
while (hts_getline(fp, KS_SEP_LINE, s) >= 0) {
if (s->l == 0) continue;
if (s->s[0] != '#') {
if (hts_verbose >= 2)
fprintf(stderr, "[E::%s] no sample line\n", __func__);
free(txt.s);
bcf_hdr_destroy(h);
return 0;
}
if (s->s[1] != '#' && fp->fn_aux) { // insert contigs here
int dret;
gzFile f;
kstream_t *ks;
kstring_t tmp;
tmp.l = tmp.m = 0; tmp.s = 0;
f = gzopen(fp->fn_aux, "r");
ks = ks_init(f);
while (ks_getuntil(ks, 0, &tmp, &dret) >= 0) {
int c;
kputs("##contig=<ID=", &txt); kputs(tmp.s, &txt);
ks_getuntil(ks, 0, &tmp, &dret);
kputs(",length=", &txt); kputw(atol(tmp.s), &txt);
kputsn(">\n", 2, &txt);
if (dret != '\n')
while ((c = ks_getc(ks)) != '\n' && c != -1); // skip the rest of the line
}
free(tmp.s);
ks_destroy(ks);
gzclose(f);
}
kputsn(s->s, s->l, &txt);
if (s->s[1] != '#') break;
kputc('\n', &txt);
}
h->l_text = txt.l + 1; // including NULL
h->text = txt.s;
bcf_hdr_parse(h);
// check tabix index, are all contigs listed in the header? add the missing ones
tbx_t *idx = tbx_index_load(fp->fn);
if ( idx )
{
int i, n, need_sync = 0;
const char **names = tbx_seqnames(idx, &n);
for (i=0; i<n; i++)
{
bcf_hrec_t *hrec = bcf_hdr_get_hrec(h, BCF_DT_CTG, (char*) names[i]);
if ( hrec ) continue;
hrec = (bcf_hrec_t*) calloc(1,sizeof(bcf_hrec_t));
hrec->key = strdup("contig");
bcf_hrec_add_key(hrec, "ID", strlen("ID"));
bcf_hrec_set_val(hrec, hrec->nkeys-1, (char*) names[i], strlen(names[i]), 0);
bcf_hrec_add_key(hrec, "length", strlen("length"));
bcf_hrec_set_val(hrec, hrec->nkeys-1, "-1", strlen("-1"), 0); // what is a good default value?
bcf_hdr_add_hrec(h, hrec);
need_sync = 1;
}
free(names);
tbx_destroy(idx);
if ( need_sync )
{
bcf_hdr_sync(h);
bcf_hdr_fmt_text(h);
}
}
return h;
} else return bcf_hdr_read((BGZF*)fp->fp);
}
int bcf_sr_add_reader(readers_t *files, const char *fname)
{
files->readers = (reader_t*) realloc(files->readers, sizeof(reader_t)*(files->nreaders+1));
reader_t *reader = &files->readers[files->nreaders++];
memset(reader,0,sizeof(reader_t));
int type = file_type(fname);
if ( type==IS_VCF_GZ )
{
reader->tbx = tbx_index_load(fname);
if ( !reader->tbx )
{
fprintf(stderr,"[add_reader] Could not load the index of %s\n", fname);
return 0;
}
// This is just to read the header
htsFile *file = hts_open(fname, "r", NULL);
if ( !file ) return 0;
reader->header = vcf_hdr_read(file);
hts_close(file);
// The VCF opened in binary tabix mode
reader->file = hts_open(fname, "rb", NULL);
if ( !reader->file ) return 0;
}
else if ( type==IS_BCF )
{
reader->file = hts_open(fname, "rb", NULL);
if ( !reader->file ) return 0;
reader->header = vcf_hdr_read(reader->file);
reader->bcf = bcf_index_load(fname);
if ( !reader->bcf )
{
fprintf(stderr,"[add_reader] Could not load the index of %s\n", fname);
return 0; // not indexed..?
}
}
else
{
fprintf(stderr,"Expected .vcf.gz or .bcf file\n");
return 0;
}
reader->fname = fname;
reader->filter_id = -1;
if ( files->apply_filters )
reader->filter_id = bcf_id2int(reader->header, BCF_DT_ID, "PASS");
// Update list of chromosomes
if ( files->region )
{
if ( !files->seqs )
{
files->mseqs = files->nseqs = 1;
files->seqs = (const char**) malloc(sizeof(const char*));
files->seqs[0] = files->region;
}
}
else
{
int n,i,j;
const char **names = bcf_seqnames(reader->header, &n);
for (i=0; i<n; i++)
{
for (j=0; j<files->nseqs; j++)
if ( !strcmp(names[i],files->seqs[j]) ) break;
if ( j<files->nseqs ) continue; // already have this chr
files->mseqs += 30;
files->seqs = (const char**) realloc(files->seqs, sizeof(const char*)*files->mseqs);
files->seqs[files->nseqs++] = names[i];
}
free(names);
}
files->iseq = -1;
return 1;
}
BCFOrderedReader::BCFOrderedReader(std::string file_name, std::vector<GenomeInterval>& intervals)
{
this->file_name = (file_name=="+")? "-" : file_name;
file = NULL;
hdr = NULL;
idx = NULL;
tbx = NULL;
itr = NULL;
this->intervals = intervals;
interval_index = 0;
index_loaded = false;
file = hts_open(this->file_name.c_str(), "r");
if (!file)
{
fprintf(stderr, "[%s:%d %s] Cannot open %s\n", __FILE__, __LINE__, __FUNCTION__, file_name.c_str());
exit(1);
}
ftype = file->format;
if (ftype.format!=vcf && ftype.format!=bcf)
{
fprintf(stderr, "[%s:%d %s] Not a VCF/BCF file: %s\n", __FILE__, __LINE__, __FUNCTION__, file_name.c_str());
exit(1);
}
s = {0, 0, 0};
if (file==NULL) exit(1);
hdr = bcf_alt_hdr_read(file);
if (!hdr) exit(1);
intervals_present = intervals.size()!=0;
if (ftype.format==bcf)
{
if ((idx = bcf_index_load(file_name.c_str())))
{
index_loaded = true;
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] index cannot be loaded for %s for random access, ignoring specified intervals and reading from start.\n", __FUNCTION__, file_name.c_str());
// exit(1);
}
}
}
else if (ftype.format==vcf)
{
if (ftype.compression==bgzf)
{
if ((tbx = tbx_index_load(file_name.c_str())))
{
index_loaded = true;
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] index cannot be loaded for %s for random access, ignoring specified intervals and reading from start.\n", __FUNCTION__, file_name.c_str());
// exit(1);
}
}
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] no random access support for VCF file: %s\n", __FUNCTION__, file_name.c_str());
// exit(1);
}
}
}
random_access_enabled = intervals_present && index_loaded;
};
int bcf_sr_add_reader(bcf_srs_t *files, const char *fname)
{
files->has_line = (int*) realloc(files->has_line, sizeof(int)*(files->nreaders+1));
files->has_line[files->nreaders] = 0;
files->readers = (bcf_sr_t*) realloc(files->readers, sizeof(bcf_sr_t)*(files->nreaders+1));
bcf_sr_t *reader = &files->readers[files->nreaders++];
memset(reader,0,sizeof(bcf_sr_t));
reader->file = hts_open(fname, "r");
if ( !reader->file ) return 0;
reader->type = reader->file->is_bin? FT_BCF : FT_VCF;
if (reader->file->is_compressed) reader->type |= FT_GZ;
if ( files->require_index )
{
if ( reader->type==FT_VCF_GZ )
{
reader->tbx_idx = tbx_index_load(fname);
if ( !reader->tbx_idx )
{
fprintf(stderr,"[add_reader] Could not load the index of %s\n", fname);
return 0;
}
reader->header = bcf_hdr_read(reader->file);
}
else if ( reader->type==FT_BCF_GZ )
{
reader->header = bcf_hdr_read(reader->file);
reader->bcf_idx = bcf_index_load(fname);
if ( !reader->bcf_idx )
{
fprintf(stderr,"[add_reader] Could not load the index of %s\n", fname);
return 0; // not indexed..?
}
}
else
{
fprintf(stderr,"Index required, expected .vcf.gz or .bcf file: %s\n", fname);
return 0;
}
}
else
{
if ( reader->type & FT_BCF )
{
reader->header = bcf_hdr_read(reader->file);
}
else if ( reader->type & FT_VCF )
{
reader->header = bcf_hdr_read(reader->file);
}
else
{
fprintf(stderr,"File type not recognised: %s\n", fname);
return 0;
}
files->streaming = 1;
}
if ( files->streaming && files->nreaders>1 )
{
fprintf(stderr,"[%s:%d %s] Error: %d readers, yet require_index not set\n", __FILE__,__LINE__,__FUNCTION__,files->nreaders);
return 0;
}
if ( files->streaming && files->regions )
{
fprintf(stderr,"[%s:%d %s] Error: cannot tabix-jump in streaming mode\n", __FILE__,__LINE__,__FUNCTION__);
return 0;
}
if ( !reader->header ) return 0;
reader->fname = fname;
if ( files->apply_filters )
reader->filter_ids = init_filters(reader->header, files->apply_filters, &reader->nfilter_ids);
// Update list of chromosomes
if ( !files->explicit_regs && !files->streaming )
{
int n,i;
const char **names = reader->tbx_idx ? tbx_seqnames(reader->tbx_idx, &n) : bcf_hdr_seqnames(reader->header, &n);
for (i=0; i<n; i++)
{
if ( !files->regions )
files->regions = _regions_init_string(names[i]);
else
_regions_add(files->regions, names[i], -1, -1);
}
free(names);
}
return 1;
}
int convert(int argc, char **argv)
{
if (argc < 2) return convert_help();
int c;
char *in=NULL, *out=NULL, *bim=NULL, *vid=NULL, *tmp_dir=NULL, *ped=NULL;
uint32_t num_fields, num_records, col = 2;
int i_is_set = 0,
o_is_set = 0,
f_is_set = 0,
b_is_set = 0,
v_is_set = 0,
t_is_set = 0,
p_is_set = 0,
r_is_set = 0;
while((c = getopt (argc, argv, "hi:o:f:r:b:v:t:p:c:")) != -1) {
switch (c) {
case 'c':
col = atoi(optarg);
break;
case 'p':
p_is_set = 1;
ped = optarg;
break;
case 't':
t_is_set = 1;
tmp_dir = optarg;
break;
case 'v':
v_is_set = 1;
vid = optarg;
break;
case 'b':
b_is_set = 1;
bim = optarg;
break;
case 'i':
i_is_set = 1;
in = optarg;
break;
case 'o':
o_is_set = 1;
out = optarg;
break;
case 'f':
f_is_set = 1;
num_fields = atoi(optarg);
break;
case 'r':
r_is_set = 1;
num_records = atoi(optarg);
break;
case 'h':
convert_help();
return 1;
case '?':
if ( (optopt == 'i') ||
(optopt == 'f') ||
(optopt == 'r') ||
(optopt == 't') ||
(optopt == 's') ||
(optopt == 'p') ||
(optopt == 'c') ||
(optopt == 'o') )
fprintf (stderr, "Option -%c requires an argument.\n",
optopt);
else if (isprint (optopt))
fprintf (stderr, "Unknown option `-%c'.\n", optopt);
else
fprintf (stderr, "Unknown option character `\\x%x'.\n", optopt);
default:
convert_help();
return 1;
}
}
char *type = argv[0];
if (i_is_set == 0) {
printf("Input file is not set\n");
return convert_help();
}
if (strcmp(type, "bcf") == 0) {
if ( (f_is_set == 0) || (r_is_set == 0) ) {
fprintf(stderr,"Attempting to autodetect num of records "
"and fields from %s\n", in);
//Try and auto detect the sizes, need the index
tbx_t *tbx = NULL;
hts_idx_t *idx = NULL;
htsFile *fp = hts_open(in,"rb");
if ( !fp ) {
fprintf(stderr,"Could not read %s\n", in);
return 1;
}
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) {
fprintf(stderr,"Could not read the header: %s\n", in);
return 1;
}
if (hts_get_format(fp)->format==vcf) {
tbx = tbx_index_load(in);
if ( !tbx ) {
fprintf(stderr,"Could not load TBI index: %s\n", in);
return 1;
}
} else if ( hts_get_format(fp)->format==bcf ) {
idx = bcf_index_load(in);
if ( !idx ) {
fprintf(stderr,"Could not load CSI index: %s\n", in);
return 1;
}
} else {
fprintf(stderr,
"Could not detect the file type as VCF or BCF: %s\n",
in);
return 1;
}
num_fields = hdr->n[BCF_DT_SAMPLE];
num_records = 0;
const char **seq;
int nseq;
seq = tbx ? tbx_seqnames(tbx, &nseq) :
bcf_index_seqnames(idx, hdr, &nseq);
int i;
uint32_t sum = 0;
for (i = 0; i < nseq; ++i) {
uint64_t records, v;
hts_idx_get_stat(tbx ? tbx->idx: idx, i, &records, &v);
num_records += records;
}
fprintf(stderr, "Number of records:%u\tNumber of fields:%u\n",
num_records, num_fields);
free(seq);
hts_close(fp);
bcf_hdr_destroy(hdr);
if (idx)
hts_idx_destroy(idx);
if (tbx)
tbx_destroy(tbx);
}
if (o_is_set == 0) {
out = (char*)malloc(strlen(in) + 5); // 5 for ext and \0
strcpy(out,in);
strcat(out, ".gqt");
}
if (b_is_set == 0) {
bim = (char*)malloc(strlen(in) + 5); // 5 for ext and \0
strcpy(bim,in);
strcat(bim, ".bim");
}
if (v_is_set == 0) {
vid = (char*)malloc(strlen(in) + 5); // 5 for ext and \0
strcpy(vid,in);
strcat(vid, ".vid");
}
if (t_is_set == 0) {
tmp_dir = (char*)malloc(3*sizeof(char)); // "./\0"
strcpy(tmp_dir,"./");
}
int r = bcf_wahbm(in, out, bim, vid, tmp_dir, num_fields, num_records);
return r;
}
if (strcmp(type, "ped") == 0) {
if (o_is_set == 0) {
if (p_is_set == 1) {
out = (char*)malloc(strlen(ped) + 4); // 4 for ext and \0
strcpy(out,ped);
strcat(out, ".db");
} else {
out = (char*)malloc(strlen(in) + 4); // 4 for ext and \0
strcpy(out,in);
strcat(out, ".db");
}
}
fprintf(stderr, "Creating sample database %s\n", out);
return ped_ped(in, ped, col, out);
}
return convert_help();
}
int main_tabix(int argc, char *argv[])
{
int c, min_shift = -1, is_force = 0, is_all = 0;
tbx_conf_t conf = tbx_conf_gff, *conf_ptr = NULL;
while ((c = getopt(argc, argv, "0fap:s:b:e:S:c:m:")) >= 0)
if (c == '0') conf.preset |= TBX_UCSC;
else if (c == 'f') is_force = 1;
else if (c == 'a') is_all = 1;
else if (c == 'm') min_shift = atoi(optarg);
else if (c == 's') conf.sc = atoi(optarg);
else if (c == 'b') conf.bc = atoi(optarg);
else if (c == 'e') conf.ec = atoi(optarg);
else if (c == 'c') conf.meta_char = *optarg;
else if (c == 'S') conf.line_skip = atoi(optarg);
else if (c == 'p') {
if (strcmp(optarg, "gff") == 0) conf_ptr = &tbx_conf_gff;
else if (strcmp(optarg, "bed") == 0) conf_ptr = &tbx_conf_bed;
else if (strcmp(optarg, "sam") == 0) conf_ptr = &tbx_conf_sam;
else if (strcmp(optarg, "vcf") == 0) conf_ptr = &tbx_conf_vcf;
else {
fprintf(stderr, "The type '%s' not recognised\n", optarg);
return 1;
}
}
if (optind == argc) {
fprintf(stderr, "\nUsage: bcftools tabix [options] <in.gz> [reg1 [...]]\n\n");
fprintf(stderr, "Options: -p STR preset: gff, bed, sam or vcf [gff]\n");
fprintf(stderr, " -s INT column number for sequence names (suppressed by -p) [1]\n");
fprintf(stderr, " -b INT column number for region start [4]\n");
fprintf(stderr, " -e INT column number for region end (if no end, set INT to -b) [5]\n");
fprintf(stderr, " -0 specify coordinates are zero-based\n");
fprintf(stderr, " -S INT skip first INT lines [0]\n");
fprintf(stderr, " -c CHAR skip lines starting with CHAR [null]\n");
fprintf(stderr, " -a print all records\n");
fprintf(stderr, " -f force to overwrite existing index\n");
fprintf(stderr, " -m INT set the minimal interval size to 1<<INT; 0 for the old tabix index [0]\n");
fprintf(stderr, "\n");
return 1;
}
if (is_all) { // read without random access
kstring_t s;
BGZF *fp;
s.l = s.m = 0; s.s = 0;
fp = bgzf_open(argv[optind], "r");
while (bgzf_getline(fp, '\n', &s) >= 0) puts(s.s);
bgzf_close(fp);
free(s.s);
} else if (optind + 2 > argc) { // create index
if ( !conf_ptr )
{
// auto-detect file type by file name
int l = strlen(argv[optind]);
int strcasecmp(const char *s1, const char *s2);
if (l>=7 && strcasecmp(argv[optind]+l-7, ".gff.gz") == 0) conf_ptr = &tbx_conf_gff;
else if (l>=7 && strcasecmp(argv[optind]+l-7, ".bed.gz") == 0) conf_ptr = &tbx_conf_bed;
else if (l>=7 && strcasecmp(argv[optind]+l-7, ".sam.gz") == 0) conf_ptr = &tbx_conf_sam;
else if (l>=7 && strcasecmp(argv[optind]+l-7, ".vcf.gz") == 0) conf_ptr = &tbx_conf_vcf;
}
if ( conf_ptr ) conf = *conf_ptr;
if (!is_force) {
char *fn;
FILE *fp;
fn = (char*)alloca(strlen(argv[optind]) + 5);
strcat(strcpy(fn, argv[optind]), min_shift <= 0? ".tbi" : ".csi");
if ((fp = fopen(fn, "rb")) != 0) {
fclose(fp);
fprintf(stderr, "[E::%s] the index file exists; use option '-f' to overwrite\n", __func__);
return 1;
}
}
if ( tbx_index_build(argv[optind], min_shift, &conf) )
{
fprintf(stderr,"tbx_index_build failed: Is the file bgzip-compressed? Was wrong -p [type] option used?\n");
return 1;
}
} else { // read with random access
tbx_t *tbx;
BGZF *fp;
kstring_t s;
int i;
if ((tbx = tbx_index_load(argv[optind])) == 0) return 1;
if ((fp = bgzf_open(argv[optind], "r")) == 0) return 1;
s.s = 0; s.l = s.m = 0;
for (i = optind + 1; i < argc; ++i) {
hts_itr_t *itr;
if ((itr = tbx_itr_querys(tbx, argv[i])) == 0) continue;
while (tbx_bgzf_itr_next(fp, tbx, itr, &s) >= 0) puts(s.s);
tbx_itr_destroy(itr);
}
free(s.s);
bgzf_close(fp);
tbx_destroy(tbx);
}
return 0;
}
int vcf_index_stats(char *fname, int stats)
{
char *fn_out = NULL;
FILE *out;
out = fn_out ? fopen(fn_out, "w") : stdout;
const char **seq;
int i, nseq;
tbx_t *tbx = NULL;
hts_idx_t *idx = NULL;
htsFile *fp = hts_open(fname,"r");
if ( !fp ) { fprintf(stderr,"Could not read %s\n", fname); return 1; }
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) { fprintf(stderr,"Could not read the header: %s\n", fname); return 1; }
if ( hts_get_format(fp)->format==vcf )
{
tbx = tbx_index_load(fname);
if ( !tbx ) { fprintf(stderr,"Could not load TBI index: %s\n", fname); return 1; }
}
else if ( hts_get_format(fp)->format==bcf )
{
idx = bcf_index_load(fname);
if ( !idx ) { fprintf(stderr,"Could not load CSI index: %s\n", fname); return 1; }
}
else
{
fprintf(stderr,"Could not detect the file type as VCF or BCF: %s\n", fname);
return 1;
}
seq = tbx ? tbx_seqnames(tbx, &nseq) : bcf_index_seqnames(idx, hdr, &nseq);
uint64_t sum = 0;
for (i=0; i<nseq; i++)
{
uint64_t records, v;
hts_idx_get_stat(tbx ? tbx->idx : idx, i, &records, &v);
sum+=records;
if (stats&2 || !records) continue;
bcf_hrec_t *hrec = bcf_hdr_get_hrec(hdr, BCF_HL_CTG, "ID", seq[i], NULL);
int hkey = hrec ? bcf_hrec_find_key(hrec, "length") : -1;
fprintf(out,"%s\t%s\t%" PRIu64 "\n", seq[i], hkey<0?".":hrec->vals[hkey], records);
}
if (!sum)
{
// No counts found.
// Is this because index version has no stored count data, or no records?
bcf1_t *rec = bcf_init1();
if (bcf_read1(fp, hdr, rec) >= 0)
{
fprintf(stderr,"%s index of %s does not contain any count metadata. Please re-index with a newer version of bcftools or tabix.\n", tbx ? "TBI" : "CSI", fname);
return 1;
}
bcf_destroy1(rec);
}
if (stats&2) fprintf(out, "%" PRIu64 "\n", sum);
free(seq);
fclose(out);
hts_close(fp);
bcf_hdr_destroy(hdr);
if (tbx)
tbx_destroy(tbx);
if (idx)
hts_idx_destroy(idx);
return 0;
}
void union_data::scanPhenotypes(string fbed) {
int n_includedP = 0;
int n_excludedP = 0;
//Open BED file
vrb.title("Scanning phenotype data in [" + fbed + "]");
htsFile *fp = hts_open(fbed.c_str(),"r");
if (!fp) vrb.error("Cannot open file");
tbx_t * tbx = tbx_index_load(fbed.c_str());
if (!tbx) vrb.error("Cannot open index file");
//Read header
kstring_t str = {0,0,0};
if (!hts_getline(fp, KS_SEP_LINE, &str) || !str.l || str.s[0] != tbx->conf.meta_char ) vrb.error("Cannot read header line");
//Scan file
vector < string > tokens;
unsigned int linecount =0;
if (regionPhenotype.chr != "NA"){
hts_itr_t *itr = tbx_itr_querys(tbx, regionPhenotype.get().c_str());
vrb.bullet("target region [" + regionPhenotype.get() + "]");
if (!itr) vrb.error("Cannot jump to region!");
//Read data
while (tbx_itr_next(fp, tbx, itr, &str) >= 0) {
linecount ++;
if (linecount % 100000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if (phenotype_id_to_idx.count(tokens[3])) continue;
if (filter_phenotype.check(tokens[3])) {
phenotype_id.push_back(tokens[3]);
phenotype_chr.push_back(tokens[0]);
phenotype_start.push_back(atoi(tokens[1].c_str()) + 1);
phenotype_end.push_back(atoi(tokens[2].c_str()));
pair < string, int > temp (tokens[3],phenotype_id_to_idx.size());
phenotype_id_to_idx.insert(temp);
n_includedP++;
} else n_excludedP ++;
}
tbx_itr_destroy(itr);
}else{
while (hts_getline(fp, KS_SEP_LINE, &str) >= 0) {
linecount ++;
if (linecount % 100000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
if (str.l && str.s[0] != tbx->conf.meta_char) {
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if (phenotype_id_to_idx.count(tokens[3])) continue;
if (filter_phenotype.check(tokens[3])) {
phenotype_id.push_back(tokens[3]);
phenotype_chr.push_back(tokens[0]);
phenotype_start.push_back(atoi(tokens[1].c_str()) + 1);
phenotype_end.push_back(atoi(tokens[2].c_str()));
pair < string, int > temp (tokens[3],phenotype_id_to_idx.size());
phenotype_id_to_idx.insert(temp);
n_includedP++;
} else n_excludedP ++;
}
}
}
//Finalize & verbose
tbx_destroy(tbx);
if (hts_close(fp)) vrb.error("Cannot properly close file");
phenotype_count = phenotype_id.size();
vrb.bullet(stb.str(n_includedP) + " new phenotypes included");
if (n_excludedP > 0) vrb.bullet(stb.str(n_excludedP) + " phenotypes excluded by user");
if (phenotype_count == 0) vrb.leave("Cannot find phenotypes in region!");
}
void union_data::readPhenotypes(string fbed, string region) {
int n_includedS = 0;
int n_includedP = 0;
int n_excludedP = 0;
vector < int > mappingS;
phenotype_id.clear();
phenotype_chr.clear();
phenotype_start.clear();
phenotype_end.clear();
phenotype_val.clear();
phenotype_count=0;
phenotype_id_to_idx.clear();
//Open BED file
//vrb.title("Reading phenotype data in [" + fbed + "]");
htsFile *fp = hts_open(fbed.c_str(),"r");
if (!fp) vrb.error("Cannot open file");
tbx_t *tbx = tbx_index_load(fbed.c_str());
if (!tbx) vrb.error("Cannot open index file");
kstring_t str = {0,0,0};
if (hts_getline(fp, KS_SEP_LINE, &str) <= 0 || !str.l || str.s[0] != tbx->conf.meta_char ) vrb.error("Cannot read header line!");
//Process sample names
vector < string > tokens;
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
for (int t = 6 ; t < tokens.size() ; t ++) {
mappingS.push_back(findSample(tokens[t]));
if (mappingS.back() >= 0) n_includedS++;
}
unsigned int linecount =0;
//Read phenotypes
hts_itr_t *itr = tbx_itr_querys(tbx, region.c_str());
//vrb.bullet("target region [" + regionPhenotype.get() + "]");
//if (!itr) vrb.error("Cannot jump to region!");
//Read data
while (tbx_itr_next(fp, tbx, itr, &str) >= 0) {
linecount ++;
if (linecount % 100000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if (filter_phenotype.check(tokens[3])) {
phenotype_id.push_back(tokens[3]);
phenotype_chr.push_back(tokens[0]);
phenotype_start.push_back(atoi(tokens[1].c_str()) + 1);
phenotype_end.push_back(atoi(tokens[2].c_str()));
phenotype_val.push_back(vector < float > (sample_count, 0.0));
for (int t = 6 ; t < tokens.size() ; t ++) {
if (mappingS[t-6] >= 0) {
if (tokens[t] == "NA") phenotype_val.back()[mappingS[t-6]] = bcf_float_missing;
else phenotype_val.back()[mappingS[t-6]] = stof(tokens[t]);
}
}
pair < string, int > temp (tokens[3],n_includedP);
phenotype_id_to_idx.insert(temp);
n_includedP++;
} else n_excludedP ++;
}
tbx_itr_destroy(itr);
//Finalize & verbose
tbx_destroy(tbx);
if (hts_close(fp)) vrb.error("Cannot properly close file");
phenotype_count = phenotype_id.size();
//vrb.bullet(stb.str(n_includedP) + " phenotypes included");
//if (n_excludedP > 0) vrb.bullet(stb.str(n_excludedP) + " phenotypes excluded by user");
//if (phenotype_count == 0) vrb.leave("Cannot find phenotypes in target region!");
}
/**
* tabix workhorse function
*/
static int tabix_handler(request_rec *r)
{
htsFile *fp=NULL;
hts_itr_t *itr=NULL;
kstring_t line = {0,0,0};
int print_header=1;
int print_body=1;
struct tabix_callback_t handler;
int http_status=OK;
memset((void*)&handler,0,sizeof(struct tabix_callback_t));
handler.r=r;
handler.limit=DEFAULT_LIMIT_RECORDS;
if (!r->handler || strcmp(r->handler, "tabix-handler")) return (DECLINED);
if (strcmp(r->method, "GET")!=0) return DECLINED;
if(r->canonical_filename==NULL) return DECLINED;
/* file must be b-gzipped */
if( !(
str_ends_with(r->canonical_filename,".gz")
)) return DECLINED;
/* file must be indexed with tabix */
if( !(
fileExtExists(r->canonical_filename,".tbi")
)) return 404;
handler.httParams = HttpParamParseGET(r);
if(handler.httParams==NULL) return DECLINED;
handler.file_format=E_FORMAT_UNDEFINED;
if(str_ends_with(r->canonical_filename,".vcf.gz"))
{
handler.file_format=E_FORMAT_VCF;
}
else if(str_ends_with(r->canonical_filename,".bed.gz"))
{
handler.file_format=E_FORMAT_BED;
}
/* only one loop, we use this to cleanup the code, instead of using a goto statement */
do {
const char* format=HttpParamGet(handler.httParams,"format");
const char* limit=HttpParamGet(handler.httParams,"limit");
const char* region=HttpParamGet(handler.httParams,"region");
int iterator_was_requested=FALSE;
if(limit!=NULL)
{
handler.limit=atol(limit);
}
if(format==NULL)
{
http_status=DECLINED;
break;
}
else if(strcmp(format,"xml")==0)
{
SETUP_HANDLER(xml);
}
else if(strcmp(format,"json")==0 || strcmp(format,"jsonp")==0)
{
handler.jsonp_callback=HttpParamGet(handler.httParams,"callback");
SETUP_HANDLER(json);
}
else if(strcmp(format,"html")==0)
{
SETUP_HANDLER(html);
}
else
{
SETUP_HANDLER(plain);
}
fp=hts_open(r->canonical_filename,"r");
if(fp==NULL)
{
http_status=HTTP_NOT_FOUND;
break;
}
//read index
handler.tbx = tbx_index_load(r->canonical_filename);
if(handler.tbx==NULL)
{
http_status=HTTP_INTERNAL_SERVER_ERROR;
break;
}
if(region!=NULL && !str_is_empty(region))
{
iterator_was_requested=TRUE;
itr = tbx_itr_querys(handler.tbx,region);
}
handler.startdocument(&handler);
if(print_header)
{
handler.startheader(&handler);
while ( hts_getline(fp, KS_SEP_LINE, &line) >= 0 )
{
if ( !line.l || line.s[0]!=handler.tbx->conf.meta_char ) break;
handler.header(&handler,&line);
handler.count++;
}
handler.enddheader(&handler);
}
handler.count=0;//Reset
if(print_body)
{
handler.startbody(&handler);
if(iterator_was_requested)
{
if(itr!=NULL)
{
while ((handler.limit==-1 || handler.count< handler.limit) && tbx_itr_next(fp, handler.tbx, itr, &line) >= 0)
{
if(handler.show(&handler,&line)<0) break;
handler.count++;
}
}
}
else
{
while ((handler.limit==-1 || handler.count< handler.limit) && \
hts_getline(fp, KS_SEP_LINE, &line) >= 0)
{
if(handler.show(&handler,&line)<0) break;
handler.count++;
}
}
handler.endbody(&handler);
}
handler.enddocument(&handler);
} while(0);/* always abort */
//cleanup
if(itr!=NULL) tbx_itr_destroy(itr);
HttpParamFree(handler.httParams);
free(line.s);
if(fp!=NULL) hts_close(fp);
if(handler.tbx!=NULL) tbx_destroy(handler.tbx);
return http_status;
}
void union_data::readGenotypesBED(string fbed,string region) {
string buffer;
int n_includedG = 0;
int n_excludedG_user = 0;
int n_includedS = 0;
int n_excludedS = 0;
int n_missingS = 0;
vector < int > mappingS;
genotype_id.clear();
genotype_chr.clear();
genotype_start.clear();
genotype_end.clear();
genotype_val.clear();
genotype_count=0;
genotype_id_to_idx.clear();
//Opening files
htsFile *fp = hts_open(fbed.c_str(),"r");
if (!fp) vrb.error("Cannot open file!");
tbx_t * tbx = tbx_index_load(fbed.c_str());
if (!tbx) vrb.error("Cannot load index file!");
kstring_t str = {0,0,0};
if (hts_getline(fp, KS_SEP_LINE, &str) <= 0 || !str.l || str.s[0] != tbx->conf.meta_char ) vrb.error("Cannot read header line!");
//Process sample names
vector < string > tokens;
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
for (int i0 = 6 ; i0 < tokens.size() ; i0 ++) {
string sid = tokens[i0];
if (filter_sample.check(sid)) {
mappingS.push_back(findSample(sid));
if (mappingS.back() >= 0) n_includedS ++;
else n_missingS ++;
} else {
mappingS.push_back(-1);
n_excludedS ++;
}
}
//vrb.bullet(stb.str(n_includedS) + " samples included");
//if (n_excludedS > 0) vrb.bullet(stb.str(n_excludedS) + " samples excluded by user");
//if (n_missingS > 0) vrb.bullet(stb.str(n_missingS) + " samples without phenotype data");
//if (n_includedS != sample_count) vrb.error("Cannot find genotype for " + stb.str(sample_count - n_includedS) + " samples!");
unsigned int linecount = 0;
//Jump to interesting region
hts_itr_t *itr = tbx_itr_querys(tbx, region.c_str());
//vrb.bullet("target region [" + regionGenotype.get() + "]");
//if (!itr) vrb.error("Cannot jump to region!");
while (tbx_itr_next(fp, tbx, itr, &str) >= 0) {
linecount ++;
if (linecount % 100000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if (filter_genotype.check(tokens[3])) {
genotype_id.push_back(tokens[3]);
genotype_chr.push_back(tokens[0]);
genotype_start.push_back(atoi(tokens[1].c_str()) + 1);
genotype_end.push_back(atoi(tokens[2].c_str()));
genotype_val.push_back(vector < float > (sample_count, 0.0));
for (int t = 6 ; t < tokens.size() ; t ++) {
if (mappingS[t-6] >= 0) {
if (tokens[t] == "NA") genotype_val.back()[mappingS[t-6]] = bcf_float_missing;
else genotype_val.back()[mappingS[t-6]] = stof(tokens[t]);
}
}
pair < string, int > temp (tokens[3],n_includedG);
genotype_id_to_idx.insert(temp);
n_includedG++;
} else n_excludedG_user ++;
}
tbx_itr_destroy(itr);
//Finalize & verbose
tbx_destroy(tbx);
if (hts_close(fp)) vrb.error("Cannot properly close file!");
genotype_count = n_includedG;
//vrb.bullet(stb.str(n_includedG) + " variants included");
//if (n_excludedG_user > 0) vrb.bullet(stb.str(n_excludedG_user) + " variants excluded by user");
//if (genotype_count == 0) vrb.leave("Cannot find variants in target region!");
}
static int query_regions(args_t *args, char *fname, char **regs, int nregs)
{
int i;
htsFile *fp = hts_open(fname,"r");
if ( !fp ) error("Could not read %s\n", fname);
enum htsExactFormat format = hts_get_format(fp)->format;
regidx_t *reg_idx = NULL;
if ( args->targets_fname )
{
reg_idx = regidx_init(args->targets_fname, NULL, NULL, 0, NULL);
if ( !reg_idx ) error("Could not read %s\n", args->targets_fname);
}
if ( format == bcf )
{
htsFile *out = hts_open("-","w");
if ( !out ) error("Could not open stdout\n", fname);
hts_idx_t *idx = bcf_index_load(fname);
if ( !idx ) error("Could not load .csi index of %s\n", fname);
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) error("Could not read the header: %s\n", fname);
if ( args->print_header )
bcf_hdr_write(out,hdr);
if ( !args->header_only )
{
bcf1_t *rec = bcf_init();
for (i=0; i<nregs; i++)
{
hts_itr_t *itr = bcf_itr_querys(idx,hdr,regs[i]);
while ( bcf_itr_next(fp, itr, rec) >=0 )
{
if ( reg_idx && !regidx_overlap(reg_idx, bcf_seqname(hdr,rec),rec->pos,rec->pos+rec->rlen-1, NULL) ) continue;
bcf_write(out,hdr,rec);
}
tbx_itr_destroy(itr);
}
bcf_destroy(rec);
}
if ( hts_close(out) ) error("hts_close returned non-zero status for stdout\n");
bcf_hdr_destroy(hdr);
hts_idx_destroy(idx);
}
else if ( format==vcf || format==sam || format==unknown_format )
{
tbx_t *tbx = tbx_index_load(fname);
if ( !tbx ) error("Could not load .tbi/.csi index of %s\n", fname);
kstring_t str = {0,0,0};
if ( args->print_header )
{
while ( hts_getline(fp, KS_SEP_LINE, &str) >= 0 )
{
if ( !str.l || str.s[0]!=tbx->conf.meta_char ) break;
puts(str.s);
}
}
if ( !args->header_only )
{
int nseq;
const char **seq = NULL;
if ( reg_idx ) seq = tbx_seqnames(tbx, &nseq);
for (i=0; i<nregs; i++)
{
hts_itr_t *itr = tbx_itr_querys(tbx, regs[i]);
if ( !itr ) continue;
while (tbx_itr_next(fp, tbx, itr, &str) >= 0)
{
if ( reg_idx && !regidx_overlap(reg_idx,seq[itr->curr_tid],itr->curr_beg,itr->curr_end, NULL) ) continue;
puts(str.s);
}
tbx_itr_destroy(itr);
}
free(seq);
}
free(str.s);
tbx_destroy(tbx);
}
else if ( format==bam )
error("Please use \"samtools view\" for querying BAM files.\n");
if ( reg_idx ) regidx_destroy(reg_idx);
if ( hts_close(fp) ) error("hts_close returned non-zero status: %s\n", fname);
for (i=0; i<nregs; i++) free(regs[i]);
free(regs);
return 0;
}
void union_data::scanGenotypesBED(string fbed) {
string buffer;
int n_includedG = 0;
int n_excludedG_user = 0;
//Opening files
htsFile *fp = hts_open(fbed.c_str(),"r");
if (!fp) vrb.error("Cannot open file!");
tbx_t * tbx = tbx_index_load(fbed.c_str());
if (!tbx) vrb.error("Cannot load index file!");
kstring_t str = {0,0,0};
if (hts_getline(fp, KS_SEP_LINE, &str) <= 0 || !str.l || str.s[0] != tbx->conf.meta_char ) vrb.error("Cannot read header line!");
//Read genotype data
vector < string > tokens;
unsigned int linecount = 0;
//Jump to interesting region
if (regionGenotype.chr != "NA"){
hts_itr_t *itr = tbx_itr_querys(tbx, regionGenotype.get().c_str());
vrb.bullet("target region [" + regionGenotype.get() + "]");
if (!itr) vrb.error("Cannot jump to region!");
while (tbx_itr_next(fp, tbx, itr, &str) >= 0) {
linecount ++;
if (linecount % 1000000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if (genotype_id_to_idx.count(tokens[3])) continue;
if (filter_genotype.check(tokens[3])) {
genotype_id.push_back(tokens[3]);
genotype_chr.push_back(tokens[0]);
genotype_start.push_back(atoi(tokens[1].c_str()) + 1);
genotype_end.push_back(atoi(tokens[2].c_str()));
pair < string, int > temp (tokens[3],genotype_id_to_idx.size());
genotype_id_to_idx.insert(temp);
n_includedG++;
} else n_excludedG_user ++;
}
tbx_itr_destroy(itr);
}else{
while (hts_getline(fp, KS_SEP_LINE, &str) >= 0) {
linecount ++;
if (linecount % 1000000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (str.l && str.s[0] != tbx->conf.meta_char) {
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if (genotype_id_to_idx.count(tokens[3])) continue;
if (filter_genotype.check(tokens[3])) {
genotype_id.push_back(tokens[3]);
genotype_chr.push_back(tokens[0]);
genotype_start.push_back(atoi(tokens[1].c_str()) + 1);
genotype_end.push_back(atoi(tokens[2].c_str()));
pair < string, int > temp (tokens[3],genotype_id_to_idx.size());
genotype_id_to_idx.insert(temp);
n_includedG++;
} else n_excludedG_user ++;
}
}
}
//Finalize & verbose
tbx_destroy(tbx);
genotype_count += n_includedG;
if (hts_close(fp)) vrb.error("Cannot properly close file!");
vrb.bullet(stb.str(n_includedG) + " new variants included");
if (n_excludedG_user > 0) vrb.bullet(stb.str(n_excludedG_user) + " variants excluded by user");
if (n_includedG == 0) vrb.leave("Cannot find variants in target region!");
}
int beds_database_add(struct beds_options *opts, const char *fname, char *columns)
{
if ( opts->n_files == opts->m_files ) {
opts->m_files = opts->m_files == 0 ? 2 : opts->m_files +2;
opts->files = (struct beds_anno_file*)realloc(opts->files, opts->m_files*sizeof(struct beds_anno_file));
}
struct beds_anno_file *file = &opts->files[opts->n_files];
memset(file, 0, sizeof(struct beds_anno_file));
file->id = opts->n_files;
file->fname = strdup(fname);
file->fp = hts_open(fname, "r");
if (file->fp == NULL)
error("Failed to open %s : %s", fname, strerror(errno));
// int n;
file->idx = tbx_index_load(fname);
if ( file->idx == NULL)
error("Failed to load index of %s.", fname);
opts->n_files++;
file->last_id = -1;
file->last_start = -1;
file->last_end = -1;
kstring_t string = KSTRING_INIT;
int no_columns = 0;
int i;
if ( columns == NULL && file->no_such_chrom == 0) {
warnings("No columns string specified for %s. Will annotate all tags in this data.", fname);
file->no_such_chrom = 1;
no_columns = 1;
} else {
int *splits = NULL;
kputs(columns, &string);
int nfields;
splits = ksplit(&string, ',', &nfields);
file->m_cols = nfields;
file->cols = (struct anno_col*)malloc(sizeof(struct anno_col) * file->m_cols);
for ( i = 0; i < nfields; ++i ) {
char *ss = string.s + splits[i];
struct anno_col *col = &file->cols[file->n_cols];
col->icol = i;
col->replace = REPLACE_MISSING;
if (*ss == '+') {
col->replace = REPLACE_MISSING;
ss++;
} else if ( *ss == '-' ) {
col->replace = REPLACE_EXISTING;
ss++;
}
if (ss[0] == '\0')
continue;
if ( strncmp(ss, "INFO/", 5) == 0)
ss += 5;
col->hdr_key = strdup(ss);
col->icol = -1;
// debug_print("%s, %d", col->hdr_key, file->n_cols);
file->n_cols++;
}
string.l = 0;
}
while (1) {
string.l =0;
if ( hts_getline(file->fp, KS_SEP_LINE, &string) < 0 )
break;
// only accept header line in the beginning for file
if ( string.s[0] != '#' )
break;
if ( strncmp(string.s, "##INFO=", 7) == 0) {
char *ss = string.s + 11;
char *se = ss;
while (se && *se != ',') se++;
struct anno_col *col = NULL;
// if no column string specified, init all header lines
if ( no_columns ) {
if ( file->n_cols == file->m_cols ) {
file->m_cols = file->m_cols == 0 ? 2 : file->m_cols + 2;
file->cols = (struct anno_col *) realloc(file->cols, file->m_cols*sizeof(struct anno_col));
}
col = &file->cols[file->n_cols++];
col->icol = -1;
col->hdr_key = strndup(ss, se-ss+1);
col->hdr_key[se-ss] = '\0';
} else {
for ( i = 0; i < file->n_cols; ++i ) {
if ( strncmp(file->cols[i].hdr_key, ss, se-ss) == 0)
break;
}
// if header line is not set in the column string, skip
if ( i == file->n_cols )
continue;
col = &file->cols[i];
}
// specify setter functions here
col->setter.bed = beds_setter_info_string;
bcf_hdr_append(opts->hdr_out, string.s);
bcf_hdr_sync(opts->hdr_out);
int hdr_id = bcf_hdr_id2int(opts->hdr_out, BCF_DT_ID,col->hdr_key);
assert ( bcf_hdr_idinfo_exists(opts->hdr_out, BCF_HL_INFO, hdr_id) );
}
string.l = 0;
// set column number for each col
if ( strncasecmp(string.s, "#chr", 4) == 0) {
int nfields;
int *splits = ksplit(&string, '\t', &nfields);
if (nfields < 4) {
fprintf(stderr, "[error] Bad header of bed database : %s. n_fields : %d, %s", fname, nfields, string.s);
fprintf(stderr, "[notice] this error usually happened because the header line is seperated by spaces but not tab!");
exit(1);
}
int k;
for ( k = 3; k < nfields; ++k ) {
char *ss = string.s + splits[k];
for (i = 0; i < file->n_cols; ++i ) {
struct anno_col *col = &file->cols[i];
if ( strcmp(col->hdr_key, ss) == 0)
break;
}
// if name line specify more names than column string or header, skip
if ( i == file->n_cols )
continue;
struct anno_col *col = &file->cols[i];
col->icol = k;
}
}
}
for ( i = 0; i < file->n_cols; ++i ) {
struct anno_col *col = &file->cols[i];
if ( col->hdr_key && col->icol == -1 )
error("No column %s found in bed database : %s", col->hdr_key, fname);
int hdr_id = bcf_hdr_id2int(opts->hdr_out, BCF_DT_ID, col->hdr_key);
assert(hdr_id>-1);
col->number = bcf_hdr_id2length(opts->hdr_out, BCF_HL_INFO, hdr_id);
if ( col->number == BCF_VL_A || col->number == BCF_VL_R || col->number == BCF_VL_G)
error("Only support fixed INFO number for bed database. %s", col->hdr_key);
col->ifile = file->id;
}
if ( string.m )
free(string.s);
if ( opts->beds_is_inited == 0 )
opts->beds_is_inited = 1;
return 0;
}
int
main_vcfset(int argc, char *argv[])
{
vcfset_conf_t vcfset_conf;
char *vcf_header = NULL;
int rc = 0;
char *vcf_in1, *vcf_in2, *vcf_out;
long int num_vars_vcf1;
long int num_vars_vcf1_ign, num_vars_out;
static int only_passed = 0;
static int only_pos = 0;
static int only_snvs = 0;
static int only_indels = 0;
static int count_only = 0;
tbx_t *vcf2_tbx = NULL; /* index for second vcf file */
htsFile *vcf2_hts = NULL;
char *add_info_field = NULL;
int vcf_concat_findex = 0;
vcf_in1 = vcf_in2 = vcf_out = NULL;
num_vars_vcf1 = 0;
num_vars_vcf1_ign = num_vars_out = 0;
/* default vcfset options */
memset(&vcfset_conf, 0, sizeof(vcfset_conf_t));
/* vcfset_conf.vcf_in1 = NULL; */
/* vcfset_conf.vcf_in2 = NULL; */
/* vcfset_conf.vcf_out = stdout;*/
/* keep in sync with long_opts_str and usage
*
* getopt is a pain in the whole when it comes to syncing of long
* and short args and usage. check out gopt, libcfu...
*/
while (1) {
int c;
static struct option long_opts[] = {
/* see usage sync */
{"help", no_argument, NULL, 'h'},
{"verbose", no_argument, &verbose, 1},
{"debug", no_argument, &debug, 1},
{"only-passed", no_argument, &only_passed, 1},
{"only-pos", no_argument, &only_pos, 1},
{"only-indels", no_argument, &only_indels, 1},
{"only-snvs", no_argument, &only_snvs, 1},
{"count-only", no_argument, &count_only, 1},
{"vcf1", required_argument, NULL, '1'},
{"vcf2", required_argument, NULL, '2'},
{"vcfout", required_argument, NULL, 'o'},
{"action", required_argument, NULL, 'a'},
{"add-info", required_argument, NULL, 'I'},
{0, 0, 0, 0} /* sentinel */
};
/* keep in sync with long_opts and usage */
static const char *long_opts_str = "h1:2:o:a:I:";
/* getopt_long stores the option index here. */
int long_opts_index = 0;
c = getopt_long(argc-1, argv+1, /* skipping 'lofreq', just leaving 'command', i.e. call */
long_opts_str, long_opts, & long_opts_index);
if (c == -1) {
break;
}
switch (c) {
/* keep in sync with long_opts etc */
case 'h':
usage(& vcfset_conf);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 0;
case '1':
vcf_in1 = strdup(optarg);
break;
case '2':
vcf_in2 = strdup(optarg);
break;
case 'o':
if (0 != strcmp(optarg, "-")) {
if (file_exists(optarg)) {
LOG_FATAL("Cowardly refusing to overwrite file '%s'. Exiting...\n", optarg);
free(vcf_in1); free(vcf_in2);
return 1;
}
}
vcf_out = strdup(optarg);
break;
case 'a':
if (0 == strcmp(optarg, "intersect")) {
vcfset_conf.vcf_setop = SETOP_INTERSECT;
} else if (0 == strcmp(optarg, "complement")) {
vcfset_conf.vcf_setop = SETOP_COMPLEMENT;
} else if (0 == strcmp(optarg, "concat")) {
vcfset_conf.vcf_setop = SETOP_CONCAT;
} else {
LOG_FATAL("Unknown action '%s'. Exiting...\n", optarg);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
break;
case 'I':
add_info_field = strdup(optarg);
break;
case '?':
LOG_FATAL("%s\n", "unrecognized arguments found. Exiting...\n");
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
default:
break;
}
}
vcfset_conf.only_passed = only_passed;
vcfset_conf.only_pos = only_pos;
vcfset_conf.only_snvs = only_snvs;
vcfset_conf.only_indels = only_indels;
if (vcfset_conf.only_indels && vcfset_conf.only_snvs) {
LOG_FATAL("%s\n", "Can't take only indels *and* only snvs into account");
return 1;
}
if (0 != argc - optind - 1) {
if (vcfset_conf.vcf_setop == SETOP_CONCAT) {
vcf_concat_findex = optind;
} else {
LOG_FATAL("%s\n", "Unrecognized arguments found\n");
return 1;
}
} else {
if (vcfset_conf.vcf_setop == SETOP_CONCAT) {
LOG_FATAL("%s\n", "No extra files for concat given\n");
return 1;
}
}
#if 0
int i; for (i=optind+1; i<argc; i++) {
LOG_FIXME("argv[%d]=%s\n", i, argv[i]);
}
#endif
if (argc == 2) {
fprintf(stderr, "\n");
usage(& vcfset_conf);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
if (vcfset_conf.vcf_setop == SETOP_UNKNOWN) {
LOG_FATAL("%s\n", "No set operation specified");
usage(& vcfset_conf);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
if (vcf_in1 == NULL || (vcf_in2 == NULL && vcfset_conf.vcf_setop != SETOP_CONCAT)) {
LOG_FATAL("%s\n\n", "At least one vcf input file not specified");
usage(& vcfset_conf);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
if (vcf_in2 != NULL && vcfset_conf.vcf_setop == SETOP_CONCAT) {
LOG_FATAL("%s\n\n", "For concat just use the -1 option followed by all other vcf files instead of using -2");
usage(& vcfset_conf);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
if (vcf_file_open(& vcfset_conf.vcf_in1, vcf_in1,
HAS_GZIP_EXT(vcf_in1), 'r')) {
LOG_ERROR("Couldn't open %s\n", vcf_in1);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
if (vcf_in2) {
vcf2_hts = hts_open(vcf_in2, "r");
if (!vcf2_hts) {
LOG_FATAL("Couldn't load %s\n", vcf_in2);
return 1;
}
vcf2_tbx = tbx_index_load(vcf_in2);
if (!vcf2_tbx) {
LOG_FATAL("Couldn't load tabix index for %s\n", vcf_in2);
return 1;
}
}
/* vcf_out default if not set: stdout==- */
if (! vcf_out) {
vcf_out = malloc(2 * sizeof(char));
strcpy(vcf_out, "-");
}
if (! count_only) {
if (vcf_file_open(& vcfset_conf.vcf_out, vcf_out,
HAS_GZIP_EXT(vcf_out), 'w')) {
LOG_ERROR("Couldn't open %s\n", vcf_out);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
}
/* use meta-data/header of vcf_in1 for output
*/
LOG_DEBUG("Getting header from %s\n", vcf_in1);
if (0 != vcf_parse_header(&vcf_header, & vcfset_conf.vcf_in1)) {
LOG_WARN("%s\n", "vcf_parse_header() failed");
if (vcf_file_seek(& vcfset_conf.vcf_in1, 0, SEEK_SET)) {
LOG_FATAL("%s\n", "Couldn't rewind file to parse variants"
" after header parsing failed");
return -1;
}
} else {
if (! count_only) {
/* vcf_write_header would write *default* header */
vcf_write_header(& vcfset_conf.vcf_out, vcf_header);
}
free(vcf_header);
}
/* parse first vcf file
*/
LOG_DEBUG("Starting to parse variants from %s\n", vcf_in1);
while (1) {
var_t *var1 = NULL;
int rc;
int is_indel;
kstring_t var2_kstr = {0, 0, 0};
hts_itr_t *var2_itr = NULL;
char regbuf[1024];
int var2_match = 0;
vcf_new_var(&var1);
rc = vcf_parse_var(& vcfset_conf.vcf_in1, var1);
if (rc) {
free(var1);
if (vcfset_conf.vcf_setop != SETOP_CONCAT) {
break;
} else {
vcf_concat_findex++;
if (vcf_concat_findex==argc) {
break;
}
/* set vcf1 up anew and simply continue as if nothing happened
*/
vcf_file_close(& vcfset_conf.vcf_in1);
free(vcf_in1);
vcf_in1 = strdup(argv[vcf_concat_findex]);
LOG_DEBUG("updated vcf_in1 = %s\n", vcf_in1);
if (vcf_file_open(& vcfset_conf.vcf_in1, vcf_in1,
HAS_GZIP_EXT(vcf_in1), 'r')) {
LOG_ERROR("Couldn't open %s\n", vcf_in1);
free(vcf_in1); free(vcf_in2); free(vcf_out);
return 1;
}
if (0 != vcf_skip_header(& vcfset_conf.vcf_in1)) {
LOG_WARN("skip header failed for %s\n", vcf_in1);
}
continue;
}
}
is_indel = vcf_var_is_indel(var1);
if (vcfset_conf.only_snvs && is_indel) {
free(var1);
continue;
} else if (vcfset_conf.only_indels && ! is_indel) {
free(var1);
continue;
}
if (! vcfset_conf.only_pos && NULL != strchr(var1->alt, ',')) {
LOG_FATAL("%s\n", "No support for multi-allelic SNVs in vcf1");
return -1;
}
if (vcfset_conf.only_passed && ! VCF_VAR_PASSES(var1)) {
#ifdef TRACE
LOG_DEBUG("Skipping non-passing var1 %s:%d\n", var1->chrom, var1->pos);
#endif
num_vars_vcf1_ign += 1;
vcf_free_var(& var1);
continue;
}
if (add_info_field) {
vcf_var_add_to_info(var1, add_info_field);
}
num_vars_vcf1 += 1;
#ifdef TRACE
LOG_DEBUG("Got passing var1 %s:%d\n", var1->chrom, var1->pos);
#endif
if (vcfset_conf.vcf_setop == SETOP_CONCAT) {
num_vars_out += 1;
if (! count_only) {
vcf_write_var(& vcfset_conf.vcf_out, var1);
}
vcf_free_var(& var1);
/* skip comparison against vcf2 */
continue;
}
/* use index access to vcf2 */
snprintf(regbuf, 1024, "%s:%ld-%ld", var1->chrom, var1->pos+1, var1->pos+1);
var2_itr = tbx_itr_querys(vcf2_tbx, regbuf);
if (! var2_itr) {
var2_match = 0;
} else {
var2_match = 0;
while (tbx_itr_next(vcf2_hts, vcf2_tbx, var2_itr, &var2_kstr) >= 0) {
var_t *var2 = NULL;
int var2_is_indel = 0;
vcf_new_var(&var2);
rc = vcf_parse_var_from_line(var2_kstr.s, var2);
/* LOG_FIXME("%d:%s>%s looking at var2 %d:%s>%s (reg %s)\n",
var1->pos+1, var1->ref, var1->alt,
var2->pos+1, var2->ref, var2->alt, regbuf); */
if (rc) {
LOG_FATAL("%s\n", "Error while parsing variant returned from tabix");
return -1;
}
var2_is_indel = vcf_var_is_indel(var2);
/* iterator returns anything overlapping with that
* position, i.e. this also includes up/downstream
* indels, so make sure actual position matches */
if (var1->pos != var2->pos) {
var2_match = 0;
} else if (vcfset_conf.only_passed && ! VCF_VAR_PASSES(var2)) {
var2_match = 0;
} else if (vcfset_conf.only_snvs && var2_is_indel) {
var2_match = 0;
} else if (vcfset_conf.only_indels && ! var2_is_indel) {
var2_match = 0;
} else if (vcfset_conf.only_pos) {
#ifdef TRACE
LOG_DEBUG("Pos match for var2 %s:%d\n", var2->chrom, var2->pos);
#endif
var2_match = 1;
} else {
if (0==strcmp(var1->ref, var2->ref) && 0==strcmp(var1->alt, var2->alt)) {
#ifdef TRACE
LOG_DEBUG("Full match for var2 %s:%d\n", var2->chrom, var2->pos);
#endif
var2_match = 1;/* FIXME: check type as well i.e. snv vs indel */
}
}
vcf_free_var(&var2);
if (var2_match) {
break;/* no need to continue */
}
}
}
if (vcfset_conf.vcf_setop == SETOP_COMPLEMENT) {
/* relative complement : elements in A but not B */
if (!var2_match) {
num_vars_out += 1;
if (! count_only) {
vcf_write_var(& vcfset_conf.vcf_out, var1);
}
}
} else if (vcfset_conf.vcf_setop == SETOP_INTERSECT) {
if (var2_match) {
num_vars_out += 1;
if (! count_only) {
vcf_write_var(& vcfset_conf.vcf_out, var1);
}
}
} else {
LOG_FATAL("Internal error: unsupported vcf_setop %d\n", vcfset_conf.vcf_setop);
return 1;
}
vcf_free_var(& var1);
tbx_itr_destroy(var2_itr);
}/* while (1) */
vcf_file_close(& vcfset_conf.vcf_in1);
if (vcf_in2) {
hts_close(vcf2_hts);
tbx_destroy(vcf2_tbx);
}
LOG_VERBOSE("Parsed %d variants from 1st vcf file (ignoring %d non-passed of those)\n",
num_vars_vcf1 + num_vars_vcf1_ign, num_vars_vcf1_ign);
LOG_VERBOSE("Wrote %d variants to output\n",
num_vars_out);
if (! count_only) {
vcf_file_close(& vcfset_conf.vcf_out);
}
if (0==rc) {
if (count_only) {
printf("%ld\n", num_vars_out);
}
LOG_VERBOSE("%s\n", "Successful exit.");
}
free(vcf_in1);
free(vcf_in2);
free(vcf_out);
return rc;
}
int bcf_sr_add_reader(bcf_srs_t *files, const char *fname)
{
htsFile* file_ptr = hts_open(fname, "r");
if ( ! file_ptr ) {
files->errnum = open_failed;
return 0;
}
files->has_line = (int*) realloc(files->has_line, sizeof(int)*(files->nreaders+1));
files->has_line[files->nreaders] = 0;
files->readers = (bcf_sr_t*) realloc(files->readers, sizeof(bcf_sr_t)*(files->nreaders+1));
bcf_sr_t *reader = &files->readers[files->nreaders++];
memset(reader,0,sizeof(bcf_sr_t));
reader->file = file_ptr;
files->errnum = 0;
if ( files->require_index )
{
if ( reader->file->format.format==vcf )
{
if ( reader->file->format.compression!=bgzf )
{
files->errnum = not_bgzf;
return 0;
}
reader->tbx_idx = tbx_index_load(fname);
if ( !reader->tbx_idx )
{
files->errnum = idx_load_failed;
return 0;
}
reader->header = bcf_hdr_read(reader->file);
}
else if ( reader->file->format.format==bcf )
{
if ( reader->file->format.compression!=bgzf )
{
files->errnum = not_bgzf;
return 0;
}
reader->header = bcf_hdr_read(reader->file);
reader->bcf_idx = bcf_index_load(fname);
if ( !reader->bcf_idx )
{
files->errnum = idx_load_failed;
return 0;
}
}
else
{
files->errnum = file_type_error;
return 0;
}
}
else
{
if ( reader->file->format.format==bcf || reader->file->format.format==vcf )
{
reader->header = bcf_hdr_read(reader->file);
}
else
{
files->errnum = file_type_error;
return 0;
}
files->streaming = 1;
}
if ( files->streaming && files->nreaders>1 )
{
files->errnum = api_usage_error;
fprintf(stderr,"[%s:%d %s] Error: %d readers, yet require_index not set\n", __FILE__,__LINE__,__FUNCTION__,files->nreaders);
return 0;
}
if ( files->streaming && files->regions )
{
files->errnum = api_usage_error;
fprintf(stderr,"[%s:%d %s] Error: cannot tabix-jump in streaming mode\n", __FILE__,__LINE__,__FUNCTION__);
return 0;
}
if ( !reader->header )
{
files->errnum = header_error;
return 0;
}
reader->fname = fname;
if ( files->apply_filters )
reader->filter_ids = init_filters(reader->header, files->apply_filters, &reader->nfilter_ids);
// Update list of chromosomes
if ( !files->explicit_regs && !files->streaming )
{
int n,i;
const char **names = reader->tbx_idx ? tbx_seqnames(reader->tbx_idx, &n) : bcf_hdr_seqnames(reader->header, &n);
for (i=0; i<n; i++)
{
if ( !files->regions )
files->regions = _regions_init_string(names[i]);
else
_regions_add(files->regions, names[i], -1, -1);
}
free(names);
}
return 1;
}
void cis_data::readPhenotypes(string fbed) {
int n_includedS = 0;
int n_includedP = 0;
int n_excludedP = 0;
int n_negativeStrd = 0;
vector < int > mappingS;
//Open BED file
vrb.title("Reading phenotype data in [" + fbed + "]");
htsFile *fp = hts_open(fbed.c_str(),"r");
if (!fp) vrb.error("Cannot open file");
tbx_t *tbx = tbx_index_load(fbed.c_str());
if (!tbx) vrb.error("Cannot open index file");
kstring_t str = {0,0,0};
if (hts_getline(fp, KS_SEP_LINE, &str) <= 0 || !str.l || str.s[0] != tbx->conf.meta_char ) vrb.error("Cannot read header line!");
//Process sample names
vector < string > tokens;
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
for (int t = 6 ; t < tokens.size() ; t ++) {
mappingS.push_back(findSample(tokens[t]));
if (mappingS.back() >= 0) n_includedS++;
}
//Read phenotypes
unsigned int linecount =0;
//Read phenotypes
if (regionPhenotype.chr != "NA"){
hts_itr_t *itr = tbx_itr_querys(tbx, regionPhenotype.get().c_str());
vrb.bullet("target region [" + regionPhenotype.get() + "]");
if (!itr) vrb.error("Cannot jump to region!");
//Read data
while (tbx_itr_next(fp, tbx, itr, &str) >= 0) {
linecount ++;
if (linecount % 100000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if ((grp_mode == GRP_NONE && filter_phenotype.check(tokens[3])) || (grp_mode != GRP_NONE && filter_phenotype.check(tokens[4]))) {
phenotype_id.push_back(tokens[3]);
phenotype_chr.push_back(tokens[0]);
phenotype_start.push_back(atoi(tokens[1].c_str()) + 1);
phenotype_end.push_back(atoi(tokens[2].c_str()));
if (grp_mode > 0 && full_test) phenotype_grp.push_back("ALL_GENES");
if (grp_mode > 0 && !full_test) phenotype_grp.push_back(tokens[4]);
phenotype_neg.push_back(tokens[5] == "-");
if (phenotype_neg.back()) n_negativeStrd ++;
phenotype_val.push_back(vector < float > (sample_count, 0.0));
for (int t = 6 ; t < tokens.size() ; t ++) {
if (mappingS[t-6] >= 0) {
if (tokens[t] == "NA") phenotype_val.back()[mappingS[t-6]] = bcf_float_missing;
else phenotype_val.back()[mappingS[t-6]] = stof(tokens[t]);
}
}
n_includedP++;
} else n_excludedP ++;
}
tbx_itr_destroy(itr);
}else{
while (hts_getline(fp, KS_SEP_LINE, &str) >= 0) {
linecount ++;
if (linecount % 100000 == 0) vrb.bullet("Read " + stb.str(linecount) + " lines");
stb.split(string(str.s), tokens);
if (str.l && str.s[0] != tbx->conf.meta_char) {
if (tokens.size() < 7) vrb.error("Incorrect number of columns!");
if ((grp_mode == GRP_NONE && filter_phenotype.check(tokens[3])) || (grp_mode != GRP_NONE && filter_phenotype.check(tokens[4]))) {
phenotype_id.push_back(tokens[3]);
phenotype_chr.push_back(tokens[0]);
phenotype_start.push_back(atoi(tokens[1].c_str()) + 1);
phenotype_end.push_back(atoi(tokens[2].c_str()));
if (grp_mode > 0 && full_test) phenotype_grp.push_back("ALL_GENES");
if (grp_mode > 0 && !full_test) phenotype_grp.push_back(tokens[4]);
phenotype_neg.push_back(tokens[5] == "-");
if (phenotype_neg.back()) n_negativeStrd ++;
phenotype_val.push_back(vector < float > (sample_count, 0.0));
for (int t = 6 ; t < tokens.size() ; t ++) {
if (mappingS[t-6] >= 0) {
if (tokens[t] == "NA") phenotype_val.back()[mappingS[t-6]] = bcf_float_missing;
else phenotype_val.back()[mappingS[t-6]] = stof(tokens[t]);
}
}
n_includedP++;
} else n_excludedP ++;
}
}
}
//Finalize & verbose
tbx_destroy(tbx);
if (hts_close(fp)) vrb.error("Cannot properly close file");
phenotype_count = phenotype_id.size();
vrb.bullet(stb.str(n_includedP) + " phenotypes included");
if (n_excludedP > 0) vrb.bullet(stb.str(n_excludedP) + " phenotypes excluded by user");
if (n_negativeStrd > 0 ) vrb.bullet(stb.str(n_negativeStrd) + " phenotypes are on the negative strand");
if (phenotype_count == 0) vrb.leave("Cannot find phenotypes in target region!");
}
bcf_sr_regions_t *bcf_sr_regions_init(const char *regions, int is_file, int ichr, int ifrom, int ito)
{
bcf_sr_regions_t *reg;
if ( !is_file ) return _regions_init_string(regions);
reg = (bcf_sr_regions_t *) calloc(1, sizeof(bcf_sr_regions_t));
reg->start = reg->end = -1;
reg->prev_start = reg->prev_seq = -1;
reg->file = hts_open(regions, "rb");
if ( !reg->file )
{
fprintf(stderr,"[%s:%d %s] Could not open file: %s\n", __FILE__,__LINE__,__FUNCTION__,regions);
free(reg);
return NULL;
}
reg->tbx = tbx_index_load(regions);
if ( !reg->tbx )
{
int len = strlen(regions);
int is_bed = strcasecmp(".bed",regions+len-4) ? 0 : 1;
if ( !is_bed && !strcasecmp(".bed.gz",regions+len-7) ) is_bed = 1;
if ( reg->file->format.format==vcf ) ito = 1;
// read the whole file, tabix index is not present
while ( hts_getline(reg->file, KS_SEP_LINE, ®->line) > 0 )
{
char *chr, *chr_end;
int from, to, ret;
ret = _regions_parse_line(reg->line.s, ichr,ifrom,abs(ito), &chr,&chr_end,&from,&to);
if ( ret < 0 )
{
if ( ito<0 )
ret = _regions_parse_line(reg->line.s, ichr,ifrom,ifrom, &chr,&chr_end,&from,&to);
if ( ret<0 )
{
fprintf(stderr,"[%s:%d] Could not parse the file %s, using the columns %d,%d[,%d]\n", __FILE__,__LINE__,regions,ichr+1,ifrom+1,ito+1);
hts_close(reg->file); reg->file = NULL; free(reg);
return NULL;
}
}
if ( !ret ) continue;
if ( is_bed ) from++;
*chr_end = 0;
_regions_add(reg, chr, from, to);
*chr_end = '\t';
}
hts_close(reg->file); reg->file = NULL;
if ( !reg->nseqs ) { free(reg); return NULL; }
return reg;
}
reg->seq_names = (char**) tbx_seqnames(reg->tbx, ®->nseqs);
if ( !reg->seq_hash )
reg->seq_hash = khash_str2int_init();
int i;
for (i=0; i<reg->nseqs; i++)
{
khash_str2int_set(reg->seq_hash,reg->seq_names[i],i);
}
reg->fname = strdup(regions);
reg->is_bin = 1;
return reg;
}
