GtRegionMapping* gt_region_mapping_new_encseq(GtEncseq *encseq, bool matchdesc,
bool usedesc)
{
GtRegionMapping *rm;
gt_assert(encseq);
gt_assert(!(matchdesc && usedesc));
rm = gt_calloc(1, sizeof (GtRegionMapping));
rm->encseq = gt_encseq_ref(encseq);
rm->matchdesc = matchdesc;
rm->usedesc = usedesc;
return rm;
}
GtMatchIterator* gt_match_iterator_sw_new(GtEncseq *es1, GtEncseq *es2,
GtScoreFunction *sf,
unsigned long min_len,
unsigned long max_edist,
GT_UNUSED GtError *err)
{
GtMatchIterator *mi;
GtMatchIteratorSW *mis;
gt_assert(es1 && es2 && sf);
gt_error_check(err);
mi = gt_match_iterator_create(gt_match_iterator_sw_class());
mis = (GtMatchIteratorSW*) mi;
mis->pvt = gt_calloc(1, sizeof (GtMatchIteratorSWMembers));
mis->pvt->es1 = gt_encseq_ref(es1);
mis->pvt->es2 = gt_encseq_ref(es2);
mis->pvt->sf = gt_score_function_ref(sf);
mis->pvt->min_len = min_len;
mis->pvt->max_edist = max_edist;
mis->pvt->firstali = true;
return mi;
}
GtSeqCol* gt_encseq_col_new(GtEncseq *encseq, GtError *err)
{
GtSeqCol *sc;
GtEncseqCol *esc;
gt_error_check(err);
gt_assert(encseq);
if (!gt_encseq_has_md5_support(encseq)) {
gt_error_set(err, "encoded sequence has no MD5 support");
return NULL;
}
sc = gt_seq_col_create(gt_encseq_col_class());
esc = gt_encseq_col_cast(sc);
esc->md5_tab = gt_encseq_get_md5_tab(encseq, err);
gt_assert(esc->md5_tab);
esc->encseq = gt_encseq_ref(encseq);
return sc;
}
GtCodonIterator* gt_codon_iterator_encseq_new_with_readmode(GtEncseq *encseq,
unsigned long startpos,
unsigned long length,
GtReadmode readmode,
GT_UNUSED GtError *err)
{
GtCodonIteratorEncseq *cie;
GtCodonIterator *ci;
gt_assert(encseq && startpos + length - 1 < gt_encseq_total_length(encseq));
gt_error_check(err);
ci = gt_codon_iterator_create(gt_codon_iterator_encseq_class());
cie = gt_codon_iterator_encseq_cast(ci);
cie->encseq = gt_encseq_ref(encseq);
cie->readmode = readmode;
ci->pvt->length = length;
ci->pvt->curpos = 0;
ci->pvt->startpos = startpos;
return ci;
}
GtWtree* gt_wtree_encseq_new(GtEncseq *encseq)
{
/* sample rate for compressd bitseq */
const unsigned int samplerate = 32U;
GtWtree *wtree;
GtWtreeEncseq *wtree_encseq;
wtree = gt_wtree_create(gt_wtree_encseq_class());
wtree_encseq = gt_wtree_encseq_cast(wtree);
wtree_encseq->encseq = gt_encseq_ref(encseq);
wtree_encseq->alpha = gt_alphabet_ref(gt_encseq_alphabet(encseq));
/* encoded chars + WC given by gt_alphabet_size,
we have to encode UNDEFCHAR and SEPARATOR too */
wtree_encseq->alpha_size = gt_alphabet_size(wtree_encseq->alpha) + 2;
wtree->members->num_of_symbols = (GtUword) wtree_encseq->alpha_size;
/* levels in tree: \lceil log_2(\sigma)\rceil */
wtree_encseq->levels =
gt_determinebitspervalue((GtUword) wtree_encseq->alpha_size);
wtree_encseq->root_fo = gt_wtree_encseq_fill_offset_new();
wtree_encseq->current_fo = wtree_encseq->root_fo;
wtree->members->length =
gt_encseq_total_length(encseq);
/* each level has number of symbols bits */
wtree_encseq->num_of_bits =
wtree_encseq->levels *
wtree->members->length;
wtree_encseq->bits_size =
wtree_encseq->num_of_bits / (sizeof (GtBitsequence) * CHAR_BIT);
if (wtree_encseq->num_of_bits % (sizeof (GtBitsequence) * CHAR_BIT) != 0)
wtree_encseq->bits_size++;
wtree_encseq->bits =
gt_calloc((size_t) wtree_encseq->bits_size, sizeof (GtBitsequence));
wtree_encseq->node_start = 0;
gt_wtree_encseq_fill_bits(wtree_encseq);
wtree_encseq->c_bits =
gt_compressed_bitsequence_new(wtree_encseq->bits,
samplerate,
wtree_encseq->num_of_bits);
gt_free(wtree_encseq->bits);
wtree_encseq->bits = NULL;
return wtree;
}
static int gt_seed_extend_runner(GT_UNUSED int argc,
GT_UNUSED const char **argv,
GT_UNUSED int parsed_args,
void *tool_arguments,
GtError *err)
{
GtSeedExtendArguments *arguments = tool_arguments;
GtEncseqLoader *encseq_loader = NULL;
GtEncseq *aencseq = NULL, *bencseq = NULL;
GtGreedyextendmatchinfo *grextinfo = NULL;
GtXdropmatchinfo *xdropinfo = NULL;
GtQuerymatchoutoptions *querymatchoutopt = NULL;
GtTimer *seedextendtimer = NULL;
GtExtendCharAccess cam = GT_EXTEND_CHAR_ACCESS_ANY;
GtUword errorpercentage = 0UL;
int had_err = 0;
gt_error_check(err);
gt_assert(arguments != NULL);
gt_assert(arguments->se_minidentity >= GT_EXTEND_MIN_IDENTITY_PERCENTAGE &&
arguments->se_minidentity <= 100UL);
/* Calculate error percentage from minidentity */
errorpercentage = 100UL - arguments->se_minidentity;
/* Measure whole running time */
if (arguments->benchmark || arguments->verbose) {
gt_showtime_enable();
}
if (gt_showtime_enabled())
{
seedextendtimer = gt_timer_new();
gt_timer_start(seedextendtimer);
}
/* Load encseq A */
encseq_loader = gt_encseq_loader_new();
gt_encseq_loader_enable_autosupport(encseq_loader);
aencseq = gt_encseq_loader_load(encseq_loader,
gt_str_get(arguments->dbs_indexname),
err);
if (aencseq == NULL)
had_err = -1;
/* If there is a 2nd read set: Load encseq B */
if (!had_err) {
if (strcmp(gt_str_get(arguments->dbs_queryname), "") != 0) {
bencseq = gt_encseq_loader_load(encseq_loader,
gt_str_get(arguments->dbs_queryname),
err);
} else {
bencseq = gt_encseq_ref(aencseq);
}
if (bencseq == NULL) {
had_err = -1;
gt_encseq_delete(aencseq);
}
}
gt_encseq_loader_delete(encseq_loader);
/* set character access method */
if (!had_err && (gt_option_is_set(arguments->se_option_greedy) ||
gt_option_is_set(arguments->se_option_xdrop) ||
arguments->se_alignmentwidth > 0))
{
cam = gt_greedy_extend_char_access(gt_str_get
(arguments->se_char_access_mode),
err);
if ((int) cam == -1) {
had_err = -1;
gt_encseq_delete(aencseq);
gt_encseq_delete(bencseq);
}
}
/* Use bias dependent parameters, adapted from E. Myers' DALIGNER */
if (!had_err && arguments->bias_parameters) {
const GtAlphabet *alpha = gt_encseq_alphabet(aencseq);
const double bias_factor[10] = {.690, .690, .690, .690, .780,
.850, .900, .933, .966, 1.000};
if (gt_alphabet_is_dna(alpha)) {
GtUword at, cg;
at = gt_encseq_charcount(aencseq, gt_alphabet_encode(alpha, 'a'));
at += gt_encseq_charcount(aencseq, gt_alphabet_encode(alpha, 't'));
cg = gt_encseq_charcount(aencseq, gt_alphabet_encode(alpha, 'c'));
cg += gt_encseq_charcount(aencseq, gt_alphabet_encode(alpha, 'g'));
if (at + cg > 0) {
const double ratio = (double)MIN(at, cg) / (at + cg);
int bias_index = (int)MAX(0.0, (ratio + 0.025) * 20.0 - 1.0);
gt_assert(bias_index < 10);
arguments->se_maxalilendiff = 30;
arguments->se_perc_match_hist = (GtUword)(100.0 - errorpercentage *
bias_factor[bias_index]);
if (arguments->verbose) {
printf("# Base ratio = %4.2lf -> percmathistory = "GT_WU"\n",
ratio, arguments->se_perc_match_hist);
}
} else {
had_err = -1;
}
} else {
had_err = -1;
}
if (had_err) {
gt_error_set(err, "option \"-bias-parameters\" can only be applied to "
"the DNA alphabet");
gt_encseq_delete(aencseq);
gt_encseq_delete(bencseq);
}
}
/* Prepare options for greedy extension */
if (!had_err && gt_option_is_set(arguments->se_option_greedy)) {
grextinfo = gt_greedy_extend_matchinfo_new(errorpercentage,
arguments->se_maxalilendiff,
arguments->se_historysize,
arguments->se_perc_match_hist,
arguments->se_alignlength,
cam,
arguments->se_extendgreedy);
if (arguments->benchmark) {
gt_greedy_extend_matchinfo_silent_set(grextinfo);
}
}
/* Prepare options for xdrop extension */
if (!had_err && gt_option_is_set(arguments->se_option_xdrop)) {
xdropinfo = gt_xdrop_matchinfo_new(arguments->se_alignlength,
errorpercentage,
arguments->se_xdropbelowscore,
arguments->se_extendxdrop);
if (arguments->benchmark) {
gt_xdrop_matchinfo_silent_set(xdropinfo);
}
}
/* Prepare output options */
if (!had_err && (arguments->se_alignmentwidth > 0 ||
gt_option_is_set(arguments->se_option_xdrop)))
{
querymatchoutopt
= gt_querymatchoutoptions_new(arguments->se_alignmentwidth);
if (gt_option_is_set(arguments->se_option_xdrop) ||
gt_option_is_set(arguments->se_option_greedy))
{
const GtUword sensitivity = gt_option_is_set(arguments->se_option_greedy)
? arguments->se_extendgreedy : 100;
gt_querymatchoutoptions_extend(querymatchoutopt,
errorpercentage,
arguments->se_maxalilendiff,
arguments->se_historysize,
arguments->se_perc_match_hist,
cam,
sensitivity);
}
}
/* Start algorithm */
if (!had_err) {
GtDiagbandseed dbsarguments;
dbsarguments.errorpercentage = errorpercentage;
dbsarguments.userdefinedleastlength = arguments->se_alignlength;
dbsarguments.seedlength = arguments->dbs_seedlength;
dbsarguments.logdiagbandwidth = arguments->dbs_logdiagbandwidth;
dbsarguments.mincoverage = arguments->dbs_mincoverage;
dbsarguments.maxfreq = arguments->dbs_maxfreq;
dbsarguments.memlimit = arguments->dbs_memlimit;
dbsarguments.mirror = arguments->mirror;
dbsarguments.overlappingseeds = arguments->overlappingseeds;
dbsarguments.verify = arguments->dbs_verify;
dbsarguments.verbose = arguments->verbose;
dbsarguments.debug_kmer = arguments->dbs_debug_kmer;
dbsarguments.debug_seedpair = arguments->dbs_debug_seedpair;
dbsarguments.seed_display = arguments->seed_display;
dbsarguments.extendgreedyinfo = grextinfo;
dbsarguments.extendxdropinfo = xdropinfo;
dbsarguments.querymatchoutopt = querymatchoutopt;
had_err = gt_diagbandseed_run(aencseq, bencseq, &dbsarguments, err);
/* clean up */
gt_encseq_delete(aencseq);
gt_encseq_delete(bencseq);
if (gt_option_is_set(arguments->se_option_greedy)) {
gt_greedy_extend_matchinfo_delete(grextinfo);
}
if (gt_option_is_set(arguments->se_option_xdrop)) {
gt_xdrop_matchinfo_delete(xdropinfo);
}
if (arguments->se_alignmentwidth > 0 ||
gt_option_is_set(arguments->se_option_xdrop)) {
gt_querymatchoutoptions_delete(querymatchoutopt);
}
}
if (gt_showtime_enabled()) {
if (!had_err) {
char *keystring
= gt_seed_extend_params_keystring(gt_option_is_set(arguments->
se_option_greedy),
gt_option_is_set(arguments->
se_option_xdrop),
arguments->dbs_seedlength,
arguments->se_alignlength,
arguments->se_minidentity,
arguments->se_maxalilendiff,
arguments->se_perc_match_hist,
arguments->se_extendgreedy,
arguments->se_extendxdrop,
arguments->se_xdropbelowscore);
printf("# TIME seedextend-%s", keystring);
gt_free(keystring);
gt_timer_show_formatted(seedextendtimer,
" overall " GT_WD ".%06ld\n",
stdout);
}
gt_timer_delete(seedextendtimer);
}
return had_err;
}