char *gt_condenseq_basefilename(const GtCondenseq *condenseq)
{
char *basename = NULL,
*suffix_ptr;
if (condenseq->filename != NULL) {
basename = gt_basename(condenseq->filename);
if (strlen(basename) > (size_t) 1 &&
(suffix_ptr = strrchr(basename + 1, '.')) != NULL) {
/* remove suffix */
*suffix_ptr = '\0';
}
}
return basename;
}
static void default_track_selector(GtBlock *block, GtStr *result,
GT_UNUSED void *data)
{
GtGenomeNode *top;
char *basename;
gt_assert(block && result);
gt_str_reset(result);
top = (GtGenomeNode*) gt_block_get_top_level_feature(block);
/* we take the basename of the filename to have nicer output in the
generated graphic. this might lead to ``collapsed'' tracks, if two files
with different paths have the same basename. */
basename = gt_basename(gt_genome_node_get_filename(top));
gt_str_append_cstr(result, basename);
gt_free(basename);
gt_str_append_char(result, GT_FILENAME_TYPE_SEPARATOR);
gt_str_append_cstr(result, gt_block_get_type(block));
}
static int gt_encseq_encode_runner(GT_UNUSED int argc, const char **argv,
int parsed_args, GT_UNUSED void *tool_arguments,
GtError *err)
{
int had_err = 0,
i;
GtEncseqEncodeArguments *arguments =
(GtEncseqEncodeArguments*) tool_arguments;
GtStrArray *infiles;
gt_error_check(err);
infiles = gt_str_array_new();
for (i = parsed_args; i < argc; i++) {
gt_str_array_add_cstr(infiles, argv[i]);
}
if (gt_str_length(arguments->indexname) == 0UL) {
if (gt_str_array_size(infiles) > 1UL) {
gt_error_set(err,"if more than one input file is given, then "
"option -indexname is mandatory");
had_err = -1;
} else {
char *basenameptr;
basenameptr = gt_basename(gt_str_array_get(infiles, 0UL));
gt_str_set(arguments->indexname, basenameptr);
gt_free(basenameptr);
}
}
if (!had_err) {
gt_assert(gt_str_length(arguments->indexname) > 0UL);
had_err = encode_sequence_files(infiles,
arguments->eopts,
gt_str_get(arguments->indexname),
arguments->verbose,
arguments->no_esq_header,
err);
}
if (!had_err && arguments->showstats)
show_encoded_statistics(infiles, gt_str_get(arguments->indexname));
gt_str_array_delete(infiles);
return had_err;
}
static int gt_compreads_compress_arguments_check(GT_UNUSED int rest_argc,
void *tool_arguments,
GtError *err)
{
int had_err = 0;
GtCsrHcrEncodeArguments *arguments = tool_arguments;
GtSplitter *splitter = NULL;
GtStr *buffer;
gt_error_check(err);
gt_assert(arguments);
if (gt_str_array_size(arguments->files) == 0) {
gt_error_set(err, "option \"-files\" is mandatory and requires"
" at least one filename as argument!");
had_err = -1;
}
if (!had_err) {
if (gt_str_length(arguments->name) == 0) {
if (gt_str_array_size(arguments->files) > 1UL) {
gt_error_set(err, "option \"-name\" needs to be specified"
" if more than one file is given");
had_err = -1;
}
else {
GtUword i;
char *basename;
splitter = gt_splitter_new();
basename = gt_basename(gt_str_array_get(arguments->files, 0));
buffer = gt_str_new_cstr(basename);
gt_splitter_split(splitter, gt_str_get(buffer), gt_str_length(buffer),
'.');
for (i = 0; i < gt_splitter_size(splitter) - 1; i++) {
gt_str_append_cstr(arguments->name,
gt_splitter_get_token(splitter, i));
if (i < gt_splitter_size(splitter) - 2)
gt_str_append_char(arguments->name, '.');
}
gt_free(basename);
gt_splitter_delete(splitter);
gt_str_delete(buffer);
}
}
}
if (!had_err) {
char *sampling_type = gt_str_get(arguments->method);
static const char *methods[] = { "page", "regular", "none" };
if (!strcmp(methods[0], sampling_type)) {
arguments->pagewise = true;
if (arguments->srate == GT_UNDEF_UWORD)
arguments->srate = GT_SAMPLING_DEFAULT_PAGE_RATE;
else if (arguments->srate == 0) {
gt_error_set(err, "page sampling was chosen, but sampling"
" rate was set to "GT_WU"! this seems wrong.",
arguments->srate);
had_err = -1;
}
}
else if (!strcmp(methods[1], sampling_type)) {
arguments->regular = true;
if (arguments->srate == GT_UNDEF_UWORD)
arguments->srate = GT_SAMPLING_DEFAULT_REGULAR_RATE;
else if (arguments->srate == 0) {
gt_error_set(err, "regular sampling was chosen, but sampling rate "
" was set to "GT_WU"! this seems wrong.",
arguments->srate);
had_err = -1;
}
}
else if (!strcmp(methods[2], sampling_type)) {
if (arguments->srate == GT_UNDEF_UWORD)
arguments->srate = 0;
else if (arguments->srate != 0) {
gt_error_set(err, "no sampling was chosen, but sampling rate was"
" set to "GT_WU"! this seems wrong.",
arguments->srate);
had_err = -1;
}
}
else {
gt_error_set(err, "somethings wrong with the stype option");
had_err = -1;
}
}
if (!had_err) {
if (arguments->arg_range.start != GT_UNDEF_UWORD) {
if (arguments->arg_range.start <= (GtUword) UINT_MAX) {
gt_safe_assign(arguments->qrng.start, arguments->arg_range.start);
if (arguments->arg_range.end <= (GtUword) UINT_MAX)
gt_safe_assign(arguments->qrng.end, arguments->arg_range.end);
else
had_err = -1;
}
else
had_err = -1;
}
if (had_err)
gt_error_set(err, "Range for qualities: value to large! larger than %u",
UINT_MAX);
}
return had_err;
}
static int gt_compreads_decompress_runner(GT_UNUSED int argc,
GT_UNUSED const char **argv,
GT_UNUSED int parsed_args,
void *tool_arguments, GtError *err)
{
GtCsrHcrDecodeArguments *arguments = tool_arguments;
int had_err = 0;
GtAlphabet *alpha = NULL;
GtHcrDecoder *hcrd = NULL;
GtTimer *timer = NULL;
unsigned long start,
end;
gt_error_check(err);
gt_assert(arguments);
if (gt_showtime_enabled()) {
timer = gt_timer_new_with_progress_description("start");
gt_timer_start(timer);
gt_assert(timer);
}
if (gt_str_length(arguments->smap) > 0) {
alpha = gt_alphabet_new_from_file_no_suffix(gt_str_get(arguments->smap),
err);
if (!alpha)
had_err = -1;
}
else {
alpha = gt_alphabet_new_dna();
if (!alpha)
had_err = -1;
}
if (!had_err) {
if (timer != NULL)
gt_timer_show_progress(timer, "decoding", stdout);
if (gt_str_length(arguments->name) == 0) {
char *basenameptr;
basenameptr = gt_basename(gt_str_get(arguments->file));
gt_str_set(arguments->name, basenameptr);
gt_free(basenameptr);
}
hcrd = gt_hcr_decoder_new(gt_str_get(arguments->file), alpha,
arguments->descs, timer, err);
if (hcrd == NULL)
had_err = -1;
else {
if (arguments->bench != 0) {
had_err = gt_compreads_decompress_benchmark(hcrd,
arguments->bench,
timer, err);
}
else {
if (arguments->rng.start != GT_UNDEF_ULONG
&& arguments->rng.end != GT_UNDEF_ULONG) {
if (arguments->rng.start >= gt_hcr_decoder_num_of_reads(hcrd)
|| arguments->rng.end >= gt_hcr_decoder_num_of_reads(hcrd)) {
gt_error_set(err, "range %lu-%lu includes a read number exceeding "
"the total number of reads (%lu)",
arguments->rng.start,
arguments->rng.end,
gt_hcr_decoder_num_of_reads(hcrd));
had_err = -1;
}
start = arguments->rng.start;
end = arguments->rng.end;
}
else {
start = 0;
end = gt_hcr_decoder_num_of_reads(hcrd) - 1;
}
if (!had_err) {
gt_log_log("filebasename: %s", gt_str_get(arguments->name));
if (gt_hcr_decoder_decode_range(hcrd, gt_str_get(arguments->name),
start, end, timer, err)
!= 0)
had_err = -1;
}
}
}
gt_hcr_decoder_delete(hcrd);
}
gt_alphabet_delete(alpha);
if (timer != NULL) {
gt_timer_show_progress_final(timer, stdout);
gt_timer_delete(timer);
}
if (had_err)
gt_assert(gt_error_is_set(err));
return had_err;
}
static void make_sequence_region(GtHashmap *sequence_regions,
GtStr *sequenceid,
GthRegionFactory *srf,
GthInput *input,
GtUword filenum,
GtUword seqnum)
{
GtUword offset_is_defined = false;
GtRange range, descrange;
GtGenomeNode *sr = NULL;
gt_assert(sequence_regions && sequenceid && srf && input);
if (gth_input_use_substring_spec(input)) {
range.start = gth_input_genomic_substring_from(input);
range.end = gth_input_genomic_substring_to(input);
}
else {
range = gth_input_get_relative_genomic_range(input, filenum, seqnum);
}
if (srf->use_desc_ranges) {
GtStr *description = gt_str_new();
gth_input_get_genomic_description(input, description, filenum, seqnum);
if (!gt_parse_description_range(gt_str_get(description), &descrange))
offset_is_defined = true;
gt_str_delete(description);
}
if (offset_is_defined)
range = gt_range_offset(&range, descrange.start);
else
range = gt_range_offset(&range, 1); /* 1-based */
if (!gt_str_length(sequenceid) ||
(gt_cstr_table_get(srf->used_seqids, gt_str_get(sequenceid)) &&
!offset_is_defined)) {
/* sequenceid is empty or exists already (and no offset has been parsed)
-> make one up */
GtStr *seqid;
char *base;
base = gt_basename(gth_input_get_genomic_filename(input, filenum));
seqid = gt_str_new_cstr(base);
gt_free(base);
gt_str_append_char(seqid, '|');
gt_str_append_uword(seqid, seqnum + 1); /* 1-based */
seqid_store_add(srf->seqid_store, filenum, seqnum, seqid, GT_UNDEF_UWORD);
gt_assert(!gt_cstr_table_get(srf->used_seqids, gt_str_get(seqid)));
gt_cstr_table_add(srf->used_seqids, gt_str_get(seqid));
sr = gt_region_node_new(seqid_store_get(srf->seqid_store, filenum, seqnum),
range.start, range.end);
gt_hashmap_add(sequence_regions,
(void*) gt_cstr_table_get(srf->used_seqids,
gt_str_get(seqid)),
sr);
gt_str_delete(seqid);
}
else {
/* sequenceid does not exists already (or an offset has been parsed)
-> use this one */
if (!gt_cstr_table_get(srf->used_seqids, gt_str_get(sequenceid))) {
/* no sequence region with this id exists -> create one */
gt_cstr_table_add(srf->used_seqids, gt_str_get(sequenceid));
seqid_store_add(srf->seqid_store, filenum, seqnum, sequenceid,
offset_is_defined ? descrange.start : GT_UNDEF_UWORD);
sr = gt_region_node_new(seqid_store_get(srf->seqid_store, filenum,
seqnum), range.start, range.end);
gt_hashmap_add(sequence_regions,
(void*) gt_cstr_table_get(srf->used_seqids,
gt_str_get(sequenceid)),
sr);
}
else {
GtRange prev_range, new_range;
/* sequence region with this id exists already -> modify range */
sr = gt_hashmap_get(sequence_regions, gt_str_get(sequenceid));
gt_assert(sr);
prev_range = gt_genome_node_get_range(sr);
new_range = gt_range_join(&prev_range, &range);
gt_genome_node_set_range(sr, &new_range);
seqid_store_add(srf->seqid_store, filenum, seqnum, sequenceid,
offset_is_defined ? descrange.start : GT_UNDEF_UWORD);
}
}
gt_assert(sr);
}
static int gt_condenseq_compress_runner(GT_UNUSED int argc, const char **argv,
int parsed_args, void *tool_arguments,
GtError *err)
{
GtCondenseqCompressArguments *arguments = tool_arguments;
GtLogger *logger,
*kdb_logger;
FILE *kmer_fp = NULL;
int had_err = 0;
gt_error_check(err);
gt_assert(arguments);
logger = gt_logger_new(arguments->verbose, GT_LOGGER_DEFLT_PREFIX, stderr);
kdb_logger = gt_logger_new(arguments->kdb, GT_LOGGER_DEFLT_PREFIX, stderr);
if (arguments->kdb) {
kmer_fp = gt_fa_fopen("kmer_db.out", "w", err);
gt_logger_set_target(kdb_logger, kmer_fp);
}
if (gt_str_length(arguments->indexname) == 0UL) {
char *basenameptr;
basenameptr = gt_basename(argv[parsed_args]);
gt_str_set(arguments->indexname, basenameptr);
gt_free(basenameptr);
}
if (!had_err) {
GtEncseqLoader *es_l = gt_encseq_loader_new();
arguments->input_es = gt_encseq_loader_load(es_l, argv[parsed_args], err);
if (arguments->input_es == NULL)
had_err = -1;
gt_encseq_loader_delete(es_l);
}
if (!had_err) {
if (arguments->minalignlength == GT_UNDEF_UWORD)
arguments->minalignlength = arguments->initsize != GT_UNDEF_UWORD ?
arguments->initsize / (GtUword) 3UL :
GT_UNDEF_UWORD;
if (arguments->windowsize == GT_UNDEF_UINT)
arguments->windowsize = arguments->minalignlength != GT_UNDEF_UWORD ?
(unsigned int) (arguments->minalignlength / 5U) :
GT_UNDEF_UINT;
if (arguments->windowsize < 4U)
arguments->windowsize = 4U;
if (arguments->kmersize == GT_UNDEF_UINT) {
unsigned int size =
gt_alphabet_num_of_chars(gt_encseq_alphabet(arguments->input_es));
/* size^k ~= 100000 */
gt_safe_assign(arguments->kmersize,
gt_round_to_long(gt_log_base(100000.0, (double) size)));
gt_logger_log(logger, "|A|: %u, k: %u",
size, arguments->kmersize);
}
if (arguments->windowsize == GT_UNDEF_UINT) {
arguments->windowsize = 5U * arguments->kmersize;
}
if (arguments->minalignlength == GT_UNDEF_UWORD) {
arguments->minalignlength = (GtUword) (3UL * arguments->windowsize);
}
if (arguments->initsize == GT_UNDEF_UWORD) {
arguments->initsize = (GtUword) (3UL * arguments->minalignlength);
}
}
if (!had_err &&
arguments->windowsize <= arguments->kmersize) {
gt_error_set(err, "-windowsize (%u) must be larger -kmersize (%u)!",
arguments->windowsize, arguments->kmersize);
had_err = -1;
}
if (!had_err &&
arguments->minalignlength < (GtUword) arguments->windowsize) {
gt_error_set(err, "-alignlength (" GT_WU ") must be at least "
"-windowsize (%u)!", arguments->minalignlength,
arguments->windowsize);
had_err = -1;
}
if (!had_err && (arguments->initsize < arguments->minalignlength)) {
gt_error_set(err, "-initsize (" GT_WU ") must be at least "
"-alignlength (" GT_WU ")!", arguments->initsize,
arguments->minalignlength);
had_err = -1;
}
if (!had_err) {
GtCondenseqCreator *ces_c;
if (!had_err) {
ces_c = gt_condenseq_creator_new(arguments->initsize,
arguments->minalignlength,
arguments->xdrop,
&(arguments->scores),
arguments->kmersize,
arguments->windowsize,
logger,
err);
if (ces_c == NULL)
had_err = -1;
}
if (!had_err) {
if (arguments->cutoff_value == GT_UNDEF_UWORD)
gt_condenseq_creator_use_mean_cutoff(ces_c);
else if (arguments->cutoff_value == 0)
gt_condenseq_creator_disable_cutoff(ces_c);
else
gt_condenseq_creator_set_cutoff(ces_c, arguments->cutoff_value);
gt_condenseq_creator_set_mean_fraction(ces_c, arguments->fraction);
if (arguments->prune)
gt_condenseq_creator_disable_prune(ces_c);
if (arguments->brute)
gt_condenseq_creator_enable_brute_force(ces_c);
if (!arguments->diags)
gt_condenseq_creator_disable_diagonals(ces_c);
if (arguments->full_diags)
gt_condenseq_creator_enable_full_diagonals(ces_c);
if (arguments->clean_percent != GT_UNDEF_UINT)
gt_condenseq_creator_set_diags_clean_limit(ces_c,
arguments->clean_percent);
had_err = gt_condenseq_creator_create(ces_c,
arguments->indexname,
arguments->input_es,
logger, kdb_logger, err);
gt_condenseq_creator_delete(ces_c);
}
}
gt_logger_delete(logger);
gt_logger_delete(kdb_logger);
if (arguments->kdb)
gt_fa_fclose(kmer_fp);
return had_err;
}
int gt_gtdata_show_help(const char *progname, GT_UNUSED void *unused,
GtError *err)
{
GtSplitter *splitter;
GtStr *doc_file;
lua_State *L = NULL;
char *prog, *bn;
int had_err = 0;
gt_error_check(err);
gt_assert(progname);
prog = gt_cstr_dup(progname); /* create modifiable copy for splitter */
splitter = gt_splitter_new();
gt_splitter_split(splitter, prog, strlen(prog), ' ');
doc_file = gt_get_gtdata_path(gt_splitter_get_token(splitter, 0), err);
if (!doc_file)
had_err = -1;
if (!had_err) {
gt_str_append_cstr(doc_file, "/doc/");
/* create Lua & push gtdata_doc_dir to Lua */
L = luaL_newstate();
if (!L) {
gt_error_set(err, "out of memory (cannot create new Lua state)");
had_err = -1;
}
}
if (!had_err) {
luaL_openlibs(L);
lua_pushstring(L, gt_str_get(doc_file));
lua_setglobal(L, "gtdata_doc_dir");
/* finish creating doc_file */
if (gt_splitter_size(splitter) == 1) {
/* special case for `gt` */
bn = gt_basename(progname);
gt_str_append_cstr(doc_file, bn);
gt_free(bn);
}
else {
/* general case for the tools */
gt_str_append_cstr(doc_file,
gt_splitter_get_token(splitter,
gt_splitter_size(splitter) - 1));
}
gt_str_append_cstr(doc_file, ".lua");
/* execute doc_file */
if (luaL_loadfile(L, gt_str_get(doc_file)) || lua_pcall(L, 0, 0, 0)) {
gt_error_set(err, "cannot run doc file: %s", lua_tostring(L, -1));
had_err = -1;
}
}
/* free */
if (L) lua_close(L);
gt_str_delete(doc_file);
gt_splitter_delete(splitter);
gt_free(prog);
return had_err;
}
static GtOPrval parsemkfmindex(Mkfmcallinfo *mkfmcallinfo,
int argc,
const char **argv,
GtError *err)
{
GtOptionParser *op;
GtOption *option, *optionfmout;
GtOPrval oprval;
int parsed_args;
gt_error_check(err);
mkfmcallinfo->indexnametab = gt_str_array_new();
mkfmcallinfo->outfmindex = gt_str_new();
mkfmcallinfo->leveldesc = gt_str_new();
op = gt_option_parser_new("[option ...] -ii indexfile [...]",
"Compute FM-index.");
gt_option_parser_set_mail_address(op, "<*****@*****.**>");
optionfmout = gt_option_new_string("fmout",
"specify name of FM-index to be generated\n"
"(mandatory if more than one input index "
"is specified)",
mkfmcallinfo->outfmindex, NULL);
gt_option_parser_add_option(op, optionfmout);
option = gt_option_new_filename_array("ii", "specify indices to be used",
mkfmcallinfo->indexnametab);
gt_option_is_mandatory(option);
gt_option_parser_add_option(op, option);
option = gt_option_new_string("size",
"specify size (tiny, small, medium, big)",
mkfmcallinfo->leveldesc, "medium");
gt_option_parser_add_option(op, option);
option = gt_option_new_bool("noindexpos",
"store no index positions (hence the positions of\n"
"matches in the index cannot be retrieved)",
&mkfmcallinfo->noindexpos,false);
gt_option_parser_add_option(op, option);
oprval = gt_option_parser_parse(op, &parsed_args, argc, argv, gt_versionfunc,
err);
if (oprval == GT_OPTION_PARSER_OK)
{
if (!gt_option_is_set(optionfmout))
{
if (gt_str_array_size(mkfmcallinfo->indexnametab) > 1UL)
{
gt_error_set(err,"if more than one index is given, then "
"option -fmout is mandatory");
oprval = GT_OPTION_PARSER_ERROR;
} else
{
char *basenameptr;
basenameptr = gt_basename(gt_str_array_get(mkfmcallinfo->indexnametab,
0));
gt_str_set(mkfmcallinfo->outfmindex,basenameptr);
gt_free(basenameptr);
}
}
}
gt_option_parser_delete(op);
if (oprval == GT_OPTION_PARSER_OK && parsed_args != argc)
{
gt_error_set(err,"superfluous program parameters");
oprval = GT_OPTION_PARSER_ERROR;
}
return oprval;
}
static GtOPrval parse_options(int *parsed_args,
bool doesa,
Suffixeratoroptions *so,
int argc, const char **argv, GtError *err)
{
GtOptionParser *op;
GtOption *option,
*optionshowprogress,
*optiongenomediff,
*optionii;
GtOPrval oprval;
gt_error_check(err);
op = gt_option_parser_new("[option ...] (-db file [...] | -ii index)",
doesa ? "Compute enhanced suffix array."
: "Compute packed index.");
gt_option_parser_set_mail_address(op, "<*****@*****.**>");
/* input info */
so->indexname = gt_str_new();
so->inputindex = gt_str_new();
so->db = gt_str_array_new();
/* register options for encoded sequence handling */
so->encopts = gt_encseq_options_register_encoding(op, so->indexname, so->db);
so->loadopts = gt_encseq_options_register_loading(op, so->indexname);
/* register options for index handling */
if (doesa)
so->idxopts = gt_index_options_register_esa(op, so->encopts);
else
so->idxopts = gt_index_options_register_packedidx(op, so->indexname,
so->encopts);
/* verbosity */
option = gt_option_new_verbose(&so->beverbose);
gt_option_parser_add_option(op, option);
optionshowprogress = gt_option_new_bool("showprogress",
"show a progress bar",
&so->showprogress,
false);
gt_option_parser_add_option(op, optionshowprogress);
optionii = gt_option_new_filename("ii", "specify existing encoded sequence",
so->inputindex);
gt_option_parser_add_option(op, optionii);
gt_option_is_mandatory_either(gt_encseq_options_db_option(so->encopts),
optionii);
gt_option_exclude(gt_encseq_options_db_option(so->encopts), optionii);
gt_option_exclude(optionii, gt_encseq_options_smap_option(so->encopts));
gt_option_exclude(optionii, gt_encseq_options_dna_option(so->encopts));
gt_option_exclude(optionii, gt_encseq_options_protein_option(so->encopts));
gt_option_exclude(optionii, gt_encseq_options_plain_option(so->encopts));
gt_option_exclude(optionii, gt_encseq_options_sat_option(so->encopts));
optiongenomediff = gt_option_new_bool("genomediff",
"directly process the lcp intervals using "
"the genomediff algorithm (suffix array and "
"lcp-tables are not output)",
&so->genomediff,
false);
gt_option_is_extended_option(optiongenomediff);
if (gt_index_options_outsuftab_option(so->idxopts) != NULL) {
gt_option_exclude(optiongenomediff,
gt_index_options_outsuftab_option(so->idxopts));
}
gt_option_parser_add_option(op, optiongenomediff);
/* suffixerator and friends do not take arguments */
gt_option_parser_set_min_max_args(op, 0U, 0U);
oprval = gt_option_parser_parse(op, parsed_args, argc, argv, gt_versionfunc,
err);
if (gt_str_length(so->indexname) == 0UL) {
/* we do not have an indexname yet, so there was none given in the
-indexname option and it could not be derived from the input filenames.
So it must be in the -ii parameter. */
char *basenameptr;
basenameptr = gt_basename(gt_str_get(so->inputindex));
gt_str_set(so->indexname, basenameptr);
gt_free(basenameptr);
}
gt_option_parser_delete(op);
return oprval;
}
static int gt_condenser_search_runner(GT_UNUSED int argc,
GT_UNUSED const char **argv,
GT_UNUSED int parsed_args,
void *tool_arguments,
GtError *err)
{
GtCondenserSearchArguments *arguments = tool_arguments;
int i, had_err = 0;
char *querypath = gt_str_get(arguments->querypath);
GtStr* coarse_fname = gt_str_new_cstr("coarse_");
char *db_basename = NULL;
char *suffix_ptr = NULL;
GtTimer *timer = NULL;
GtLogger *logger = NULL;
gt_error_check(err);
gt_assert(arguments);
logger = gt_logger_new(arguments->verbose, GT_LOGGER_DEFLT_PREFIX, stderr);
db_basename = gt_basename(gt_str_get(arguments->dbpath));
/* if first char is '.' this might be a hidden file */
if (strlen(db_basename) > (size_t) 1 &&
(suffix_ptr = strrchr(db_basename + 1, '.')) != NULL) {
/* remove suffix */
*suffix_ptr = '\0';
}
gt_str_append_cstr(coarse_fname, db_basename);
gt_str_append_cstr(coarse_fname, ".fas");
gt_free(db_basename);
db_basename = NULL;
suffix_ptr = NULL;
if (arguments->blastn || arguments->blastp) {
GtMatch *match;
GtMatchIterator *mp = NULL;
GtNREncseq *nrencseq = NULL;
GtStr *fastaname = gt_str_clone(arguments->dbpath);
HitPosition *hits;
double eval,
raw_eval = 0.0;
GtUword coarse_db_len = 0;
GtMatchIteratorStatus status;
int curr_hits = 0,
max_hits = 100;
hits = gt_malloc(sizeof (*hits) * (size_t) max_hits);
gt_str_append_cstr(fastaname, ".fas");
for (i=0; i < max_hits; i++) {
hits[i].range = gt_malloc(sizeof (*hits[i].range) * (size_t) 1);
}
if (gt_showtime_enabled()) {
timer = gt_timer_new_with_progress_description("initialization");
gt_timer_start(timer);
}
/*extract sequences from compressed database*/
if (!had_err) {
nrencseq = gt_n_r_encseq_new_from_file(gt_str_get(arguments->dbpath),
logger, err);
if (nrencseq == NULL)
had_err = -1;
}
if (!had_err) {
if (arguments->ceval == GT_UNDEF_DOUBLE ||
arguments->feval == GT_UNDEF_DOUBLE) {
/* from NCBI BLAST tutorial:
E = Kmne^{-lambdaS}
calculates E-value for score S with natural scale parameters K for
search space size and lambda for the scoring system
E = mn2^-S'
m being the subject (total) length, n the length of ONE query
calculates E-value for bit-score S'
*/
GtFastaReader *reader;
GtCondenserSearchAvg avg = {0,0};
reader = gt_fasta_reader_rec_new(arguments->querypath);
had_err = gt_fasta_reader_run(reader, NULL, NULL,
gt_condenser_search_cum_moving_avg,
&avg,
err);
if (!had_err) {
GtUword S = arguments->bitscore;
gt_log_log(GT_WU " queries, avg query size: " GT_WU,
avg.count, avg.avg);
raw_eval = 1/pow(2.0, (double) S) * avg.avg;
gt_logger_log(logger, "Raw E-value set to %.4e", raw_eval);
gt_assert(avg.avg != 0);
}
gt_fasta_reader_delete(reader);
}
}
/*create BLAST database from compressed database fasta file*/
if (!had_err) {
if (timer != NULL)
gt_timer_show_progress(timer, "create coarse BLAST db", stderr);
if (arguments->blastn)
had_err = gt_condenser_search_create_nucl_blastdb(gt_str_get(fastaname),
err);
else
had_err = gt_condenser_search_create_prot_blastdb(gt_str_get(fastaname),
err);
}
if (!had_err) {
GtBlastProcessCall *call;
if (timer != NULL)
gt_timer_show_progress(timer, "coarse BLAST run", stderr);
if (arguments->blastp)
call = gt_blast_process_call_new_prot();
else
call = gt_blast_process_call_new_nucl();
gt_blast_process_call_set_db(call, gt_str_get(fastaname));
gt_blast_process_call_set_query(call, querypath);
gt_blast_process_call_set_evalue(call, arguments->ceval);
gt_blast_process_call_set_num_threads(call, arguments->blthreads);
mp = gt_match_iterator_blast_process_new(call, err);
if (!mp)
had_err = -1;
gt_blast_process_call_delete(call);
while (!had_err &&
(status = gt_match_iterator_next(mp, &match, err)) !=
GT_MATCHER_STATUS_END)
{
if (status == GT_MATCHER_STATUS_OK) {
GtUword hit_seq_id;
char string[7];
const char *dbseqid = gt_match_get_seqid2(match);
if (sscanf(dbseqid,"%6s" GT_WU, string, &hit_seq_id) == 2) {
gt_match_get_range_seq2(match, hits[curr_hits].range);
hits[curr_hits].idx = hit_seq_id;
gt_match_delete(match);
curr_hits++;
if (curr_hits == max_hits) {
HitPosition *hit_extention;
max_hits += 100;
hits = gt_realloc(hits, sizeof (*hit_extention) * max_hits);
for (i=max_hits - 100; i < max_hits; i++) {
hits[i].range = gt_malloc(sizeof (*hits[i].range));
}
}
} else {
gt_error_set(err, "could not parse unique db header %s", dbseqid);
had_err = -1;
}
} else if (status == GT_MATCHER_STATUS_ERROR) {
had_err = -1;
}
}
gt_match_iterator_delete(mp);
}
/*extract sequences*/
if (!had_err) {
GtNREncseqDecompressor *decomp;
GtFile *coarse_hits;
if (timer != NULL)
gt_timer_show_progress(timer, "extract coarse search hits", stderr);
decomp = gt_n_r_encseq_decompressor_new(nrencseq);
coarse_hits = gt_file_new(gt_str_get(coarse_fname),"w", err);
/* TODO DW do NOT extract complete uniques! these could be complete
chromosomes!! just extract something around it? maybe +- max query
length*/
for (i = 0; i < curr_hits; i++) {
gt_n_r_encseq_decompressor_add_unique_idx_to_extract(decomp,
hits[i].idx);
}
had_err =
gt_n_r_encseq_decompressor_start_unique_extraction(coarse_hits,
decomp,
&coarse_db_len,
err);
gt_assert(coarse_db_len != 0);
gt_file_delete(coarse_hits);
gt_n_r_encseq_decompressor_delete(decomp);
}
gt_n_r_encseq_delete(nrencseq);
/* create BLAST database from decompressed database file */
if (!had_err) {
if (timer != NULL)
gt_timer_show_progress(timer, "create fine BLAST db", stderr);
if (arguments->blastn)
had_err =
gt_condenser_search_create_nucl_blastdb(gt_str_get(coarse_fname),
err);
else
had_err =
gt_condenser_search_create_prot_blastdb(gt_str_get(coarse_fname),
err);
}
/* perform fine BLAST search */
if (!had_err) {
GtBlastProcessCall *call;
if (timer != NULL)
gt_timer_show_progress(timer, "fine BLAST run", stderr);
if (arguments->feval == GT_UNDEF_DOUBLE) {
eval = raw_eval * coarse_db_len;
} else {
eval = arguments->feval;
}
if (arguments->blastp)
call = gt_blast_process_call_new_prot();
else
call = gt_blast_process_call_new_nucl();
gt_blast_process_call_set_db(call, gt_str_get(coarse_fname));
gt_blast_process_call_set_query(call, querypath);
gt_blast_process_call_set_evalue(call, eval);
gt_blast_process_call_set_num_threads(call, arguments->blthreads);
gt_logger_log(logger, "Fine E-value set to: %.4e (len)" GT_WU, eval,
coarse_db_len);
mp = gt_match_iterator_blast_process_new(call, err);
if (!mp)
had_err = -1;
gt_blast_process_call_delete(call);
if (!had_err) {
GtUword numofhits = 0;
while (!had_err &&
(status = gt_match_iterator_next(mp, &match, err)) !=
GT_MATCHER_STATUS_END) {
if (status == GT_MATCHER_STATUS_OK) {
GtMatchBlast *matchb = (GtMatchBlast*) match;
char *dbseqid = gt_malloc(sizeof (*dbseqid) * 50);
GtRange range_seq1;
GtRange range_seq2;
numofhits++;
gt_match_get_range_seq1(match, &range_seq1);
gt_match_get_range_seq2(match, &range_seq2);
gt_file_xprintf(
arguments->outfp,
"%s\t%s\t%.2f\t" GT_WU "\t" GT_WU "\t" GT_WU "\t" GT_WU "\t"
GT_WU "\t%g\t%.3f\n",
gt_match_get_seqid1(match),
gt_match_get_seqid2(match),
gt_match_blast_get_similarity(matchb),
gt_match_blast_get_align_length(matchb),
range_seq1.start,
range_seq1.end,
range_seq2.start,
range_seq2.end,
gt_match_blast_get_evalue(matchb),
(double) gt_match_blast_get_bitscore(matchb));
gt_match_delete(match);
gt_free(dbseqid);
} else if (status == GT_MATCHER_STATUS_ERROR) {
had_err = -1;
}
}
gt_log_log(GT_WU " hits found\n", numofhits);
}
gt_match_iterator_delete(mp);
}
if (!had_err)
if (timer != NULL)
gt_timer_show_progress_final(timer, stderr);
gt_timer_delete(timer);
/*cleanup*/
for (i=0; i < max_hits; i++) {
gt_free(hits[i].range);
}
gt_free(hits);
gt_str_delete(fastaname);
}
gt_str_delete(coarse_fname);
gt_logger_delete(logger);
return had_err;
}