static void set_gff3_target_attribute(GthSA *sa, bool md5ids)
{
gt_assert(sa && !sa->gff3_target_attribute);
sa->gff3_target_attribute = gt_str_new();
if (md5ids) {
gt_assert(sa->ref_md5);
gt_str_append_cstr(sa->gff3_target_attribute, GT_MD5_SEQID_PREFIX);
gt_str_append_str(sa->gff3_target_attribute, sa->ref_md5);
gt_str_append_char(sa->gff3_target_attribute, ':');
}
gt_gff3_escape(sa->gff3_target_attribute, gt_str_get(sa->ref_id),
gt_str_length(sa->ref_id));
gt_str_append_char(sa->gff3_target_attribute, ' ');
gt_str_append_uword(sa->gff3_target_attribute,
gth_sa_referencecutoff_start(sa) + 1); /* XXX: use
reference
dpstartpos */
gt_str_append_char(sa->gff3_target_attribute, ' ');
gt_str_append_uword(sa->gff3_target_attribute,
gth_sa_ref_total_length(sa) - /* XXX */
gth_sa_referencecutoff_end(sa));
gt_str_append_char(sa->gff3_target_attribute, ' ');
if (sa->ref_strand_forward) {
gt_str_append_char(sa->gff3_target_attribute,
GT_STRAND_CHARS[GT_STRAND_FORWARD]);
}
else {
gt_str_append_char(sa->gff3_target_attribute,
GT_STRAND_CHARS[GT_STRAND_REVERSE]);
}
}
static void showalignmentcutoffs(GthSA *sa, unsigned int indentlevel,
GtFile *outfp)
{
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<cutoffs>\n");
indentlevel++;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<cutoffsstart>\n");
indentlevel++;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<genomiccutoff>"GT_WU"</genomiccutoff>\n",
gth_sa_genomiccutoff_start(sa));
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<referencecutoff>"GT_WU"</referencecutoff>\n",
gth_sa_referencecutoff_start(sa));
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<eopcutoff>"GT_WU"</eopcutoff>\n",
gth_sa_eopcutoff_start(sa));
indentlevel--;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "</cutoffsstart>\n");
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<cutoffsend>\n");
indentlevel++;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<genomiccutoff>"GT_WU"</genomiccutoff>\n",
gth_sa_genomiccutoff_end(sa));
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<referencecutoff>"GT_WU"</referencecutoff>\n",
gth_sa_referencecutoff_end(sa));
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<eopcutoff>"GT_WU"</eopcutoff>\n",
gth_sa_eopcutoff_end(sa));
indentlevel--;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "</cutoffsend>\n");
indentlevel--;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "</cutoffs>\n");
}
void gth_compute_scores(GthSA *sa,
bool proteineop,
GthDPParam *dp_param,
void *dp_options_est,
const unsigned char *gen_seq_tran,
const unsigned char *ref_seq_tran,
const unsigned char *ref_seq_orig,
const GtTransTable *transtable,
unsigned long gen_dp_start,
unsigned long scoreminexonlen,
bool introncutout,
bool gs2out,
GthSplicedSeq *spliced_seq,
unsigned long ref_dp_length,
GtAlphabet *gen_alphabet,
GtAlphabet *ref_alphabet,
GthDPScoresProtein *dp_scores_protein)
{
Traversealignmentfunctions travfunctions;
Traversealignmentstate travstate;
Computebordersandscoresdata data;
GthFlt score, coverageofgenomicsegment, coverageofreferencesegment;
gt_assert(!gth_sa_num_of_exons(sa));
gt_assert(!gth_sa_num_of_introns(sa));
travfunctions.processmismatch = computescoresprocmismatch;
travfunctions.processdeletion = computescoresprocdeletion;
travfunctions.processinsertion = computebordersandscoresprocinsertion;
travfunctions.processmatch = computebordersandscoresprocmatch;
travfunctions.processintron = computebordersandscoresprocintron;
travfunctions.breakcondition = NULL;
/* additional functions for protein edit operations */
travfunctions.processintron_with_1_base_left =
computebordersandscoresprocintron;
travfunctions.processintron_with_2_bases_left =
computebordersandscoresprocintron;
travfunctions.processmismatch_with_1_gap =
computescoresprocmismatchordeletionwithgap;
travfunctions.processmismatch_with_2_gaps =
computescoresprocmismatchordeletionwithgap;
travfunctions.processdeletion_with_1_gap =
computescoresprocmismatchordeletionwithgap;
travfunctions.processdeletion_with_2_gaps =
computescoresprocmismatchordeletionwithgap;
travstate.proteineop = proteineop;
travstate.processing_intron_with_1_base_left = false;
travstate.processing_intron_with_2_bases_left = false;
travstate.alignment = gth_sa_get_editoperations(sa);
travstate.alignmentlength = gth_sa_get_editoperations_length(sa);
travstate.eopptr = travstate.alignment + travstate.alignmentlength - 1;
travstate.genomicptr = gth_sa_genomiccutoff_start(sa);
travstate.referenceptr = gth_sa_referencecutoff_start(sa);
if (travstate.alignmentlength <= 0) {
/* in this case the alignmentscore is set to 0, which leads to discarding
this alignment later */
gth_sa_set_score(sa, 0.0);
return;
}
/* editoperations contain no zero base exons */
gt_assert(gth_sa_contains_no_zero_base_exons(sa));
/* editoperations contain no leading or terminal introns or insertions */
gt_assert(containsnoleadingorterminalintronsorinsertions(travstate.alignment,
travstate
.alignmentlength,
proteineop));
/* sum of edit operations equals referencelength */
gt_assert(gt_eops_equal_referencelength(travstate.alignment,
travstate.alignmentlength,
ref_dp_length
- gth_sa_referencecutoff_start(sa)
- gth_sa_referencecutoff_end(sa),
proteineop));
data.proteineop = proteineop;
data.newexon = true;
data.newintron = true;
data.firstexon = true;
data.introncutout = introncutout;
data.gs2out = gs2out;
data.spliced_seq = spliced_seq;
data.singleexonweight = (GthFlt) 0.0;
data.maxsingleexonweight = (GthFlt) 0.0;
data.overallexonweight = (GthFlt) 0.0;
data.maxoverallexonweight = (GthFlt) 0.0;
data.cumulativelengthofscoredexons = 0;
data.exon.leftgenomicexonborder = GT_UNDEF_ULONG;
data.exon.rightgenomicexonborder = GT_UNDEF_ULONG;
data.exon.leftreferenceexonborder = GT_UNDEF_ULONG;
data.exon.rightreferenceexonborder = GT_UNDEF_ULONG;
data.exon.exonscore = GTH_UNDEF_GTHDBL;
data.intron.donorsiteprobability = GTH_UNDEF_GTHFLT;
data.intron.acceptorsiteprobability = GTH_UNDEF_GTHFLT;
data.intron.donorsitescore = GTH_UNDEF_GTHDBL;
data.intron.acceptorsitescore = GTH_UNDEF_GTHDBL;
data.sa = sa;
data.dp_param = dp_param;
data.dp_options_est = dp_options_est;
data.gen_seq_tran = gen_seq_tran;
data.ref_seq_tran = ref_seq_tran;
data.ref_seq_orig = ref_seq_orig;
data.transtable = transtable;
data.gen_dp_start = gen_dp_start;
data.scoreminexonlen = scoreminexonlen;
data.ref_dp_length = ref_dp_length;
data.gen_alphabet = gen_alphabet;
data.gen_alphabet_characters = gen_alphabet
? gt_alphabet_characters(gen_alphabet)
: NULL;
data.dp_scores_protein = dp_scores_protein;
gthtraversealignment(true, &travstate, proteineop, &data, &travfunctions);
/* this is for saving the last exon */
evalnewintronifpossible(proteineop, &data.newexon, &data.newintron, true,
data.introncutout, data.gs2out, data.spliced_seq,
&data.exon, &data.intron, &data.singleexonweight,
&data.maxsingleexonweight, &data.overallexonweight,
&data.maxoverallexonweight,
&data.cumulativelengthofscoredexons, sa, &travstate,
gen_alphabet, data.dp_param, data.dp_options_est,
data.gen_seq_tran, data.ref_seq_tran,
data.gen_dp_start, data.scoreminexonlen);
/* saving the scores for the whole alignment */
if (data.maxoverallexonweight > 0.0) {
score = data.overallexonweight / data.maxoverallexonweight;
/* XXX: the way the alignmentscore is computed, it is possible to get a
score > 1.0. Since we don't want this, we cap it */
if (score > 1.0)
score = 1.0;
}
else
score = 0.0;
gth_sa_set_score(sa, score);
gth_sa_set_cumlen_scored_exons(sa, data.cumulativelengthofscoredexons);
/* fraction of the gen_dp_length which is scored/weighted */
coverageofgenomicsegment = (GthFlt) data.cumulativelengthofscoredexons /
(GthFlt) gth_sa_gen_dp_length(sa);
/* coverage of genomic segment is valid value */
gt_assert(coverageofgenomicsegment >= 0.0 && coverageofgenomicsegment <= 1.0);
/* fraction of the referencelength which is scored/weighted */
coverageofreferencesegment = (GthFlt) data.cumulativelengthofscoredexons /
(GthFlt) ((proteineop ? GT_CODON_LENGTH : 1) *
gth_sa_ref_total_length(sa));
if (coverageofgenomicsegment > coverageofreferencesegment) {
gth_sa_set_coverage(sa, coverageofgenomicsegment);
gth_sa_set_highest_cov(sa, true);
}
else {
gth_sa_set_coverage(sa, coverageofreferencesegment);
gth_sa_set_highest_cov(sa, false);
}
/* test the assumption that the coverage is never larger then the default */
gt_assert(gth_sa_coverage(sa) <= GTH_DEFAULT_MAX_COVERAGE);
/* compute poly(A) tail position */
gth_sa_calc_polyAtailpos(sa, ref_seq_tran, ref_alphabet);
/* determined exons are forward and consecutive */
gt_assert(gth_sa_exons_are_forward_and_consecutive(sa));
}
bool gth_sas_are_equal(const GthSA *saA, const GthSA *saB)
{
Exoninfo *exoninfoA, *exoninfoB;
Introninfo *introninfoA, *introninfoB;
GtUword i;
/* compare element 0 */
if (gth_sa_alphatype(saA) != gth_sa_alphatype(saB))
return false;
/* compare element 1 */
if (gth_backtrace_path_length(saA->backtrace_path) !=
gth_backtrace_path_length(saB->backtrace_path)) {
return false;
}
for (i = 0; i < gth_backtrace_path_length(saA->backtrace_path); i++) {
if (((Editoperation*) gth_backtrace_path_get(saA->backtrace_path))[i] !=
((Editoperation*) gth_backtrace_path_get(saB->backtrace_path))[i]) {
return false;
}
}
/* element 2 has been removed (indelcount) */
/* compare element 3 */
if (gth_sa_gen_dp_length(saA) != gth_sa_gen_dp_length(saB))
return false;
/* compare element 4 */
if (saA->gen_total_length != saB->gen_total_length)
return false;
/* compare element 5 */
if (saA->gen_offset != saB->gen_offset)
return false;
/* compare element 6 */
if (gth_sa_ref_total_length(saA) != gth_sa_ref_total_length(saB))
return false;
/* compare element 7 */
if (gth_sa_gen_dp_start(saA) != gth_sa_gen_dp_start(saB))
return false;
/* element 8 has been removed (gen_dp_end) */
/* compare element 9 */
if (saA->gen_file_num != saB->gen_file_num)
return false;
/* compare element 10 */
if (saA->gen_seq_num != saB->gen_seq_num)
return false;
/* compare element 11 */
if (saA->ref_file_num != saB->ref_file_num)
return false;
/* compare element 12 */
if (saA->ref_seq_num != saB->ref_seq_num)
return false;
/* compare element 13 */
if (gt_str_cmp(saA->gen_id, saB->gen_id))
return false;
/* compare element 14 */
if (gt_str_cmp(saA->ref_id, saB->ref_id))
return false;
/* compare element 15 */
if (saA->gen_strand_forward != saB->gen_strand_forward)
return false;
/* compare element 16 */
if (saA->ref_strand_forward != saB->ref_strand_forward)
return false;
/* compare element 17 */
if (gth_sa_genomiccutoff_start(saA) != gth_sa_genomiccutoff_start(saB))
return false;
if (gth_sa_referencecutoff_start(saA) != gth_sa_referencecutoff_start(saB))
return false;
if (gth_sa_eopcutoff_start(saA) != gth_sa_eopcutoff_start(saB))
return false;
if (gth_sa_genomiccutoff_end(saA) != gth_sa_genomiccutoff_end(saB))
return false;
if (gth_sa_referencecutoff_end(saA) != gth_sa_referencecutoff_end(saB))
return false;
if (gth_sa_eopcutoff_end(saA) != gth_sa_eopcutoff_end(saB))
return false;
/* compare element 18 */
if (gt_array_size(saA->exons) != gt_array_size(saB->exons))
return false;
for (i = 0; i < gt_array_size(saA->exons); i++) {
exoninfoA = (Exoninfo*) gt_array_get(saA->exons, i);
exoninfoB = (Exoninfo*) gt_array_get(saB->exons, i);
if (exoninfoA->leftgenomicexonborder != exoninfoB->leftgenomicexonborder)
return false;
if (exoninfoA->rightgenomicexonborder != exoninfoB->rightgenomicexonborder)
return false;
if (exoninfoA->leftreferenceexonborder !=
exoninfoB->leftreferenceexonborder) {
return false;
}
if (exoninfoA->rightreferenceexonborder !=
exoninfoB->rightreferenceexonborder) {
return false;
}
if (!gt_double_equals_double(exoninfoA->exonscore, exoninfoB->exonscore)) {
return false;
}
}
/* compare element 19 */
if (gt_array_size(saA->introns) != gt_array_size(saB->introns))
return false;
for (i = 0; i < gt_array_size(saA->introns); i++) {
introninfoA = (Introninfo*) gt_array_get(saA->introns, i);
introninfoB = (Introninfo*) gt_array_get(saB->introns, i);
if (!gt_double_equals_double(introninfoA->donorsiteprobability,
introninfoB->donorsiteprobability)) {
return false;
}
if (!gt_double_equals_double(introninfoA->acceptorsiteprobability,
introninfoB->acceptorsiteprobability)) {
return false;
}
if (!gt_double_equals_double(introninfoA->donorsitescore,
introninfoB->donorsitescore)) {
return false;
}
if (!gt_double_equals_double(introninfoA->acceptorsitescore,
introninfoB->acceptorsitescore)) {
return false;
}
}
/* compare element 20 */
if (saA->polyAtailpos.start != saB->polyAtailpos.start)
return false;
if (saA->polyAtailpos.end != saB->polyAtailpos.end)
return false;
/* compare element 21 */
if (saA->alignmentscore != saB->alignmentscore)
return false;
/* compare element 22 */
if (saA->coverage != saB->coverage)
return false;
/* compare element 23 */
if (saA->genomic_cov_is_highest != saB->genomic_cov_is_highest)
return false;
/* compare element 24 */
if (saA->cumlen_scored_exons != saB->cumlen_scored_exons)
return false;
return true;
}
GtUword gth_sa_get_alignment_lines(const GthSA *sa,
unsigned char **first_line,
unsigned char **second_line,
unsigned char **third_line,
GtUword translationtable,
GthInput *input)
{
GtUword genomicstartcutoff, genomicendcutoff, genomictotalcutoff,
referencestartcutoff, referenceendcutoff, referencetotalcutoff;
GT_UNUSED bool reverse_subject_pos = false;
gt_assert(sa && first_line && second_line && third_line && input);
/* only for cosmetic reasons */
genomicstartcutoff = gth_sa_genomiccutoff_start(sa);
genomicendcutoff = gth_sa_genomiccutoff_end(sa);
genomictotalcutoff = genomicstartcutoff + genomicendcutoff;
referencestartcutoff = gth_sa_referencecutoff_start(sa);
referenceendcutoff = gth_sa_referencecutoff_end(sa);
referencetotalcutoff = referencestartcutoff + referenceendcutoff;
/* sequences */
unsigned char *gen_seq_orig, *ref_seq_orig;
GtUword cols = 0;
GthSeqCon *ref_seq_con;
/* make sure that the correct files are loaded */
gth_input_load_reference_file(input, gth_sa_ref_file_num(sa), false);
ref_seq_con = gth_input_current_ref_seq_con(input);
/* If the reverse complement of the genomic DNA is considered, this
opition is needed for correct output of the genomic sequence positions
by the function showalignmentgeneric() */
if (!gth_sa_gen_strand_forward(sa))
reverse_subject_pos = true;
/* get genomic sequence */
gen_seq_orig = (unsigned char*)
gth_input_original_genomic_sequence(input, gth_sa_gen_file_num(sa),
gth_sa_gen_strand_forward(sa))
+ gth_sa_gen_dp_start(sa);
/* get reference sequence */
if (gth_sa_ref_strand_forward(sa)) {
ref_seq_orig =
gth_seq_con_get_orig_seq(ref_seq_con, gth_sa_ref_seq_num(sa));
}
else {
ref_seq_orig =
gth_seq_con_get_orig_seq_rc(ref_seq_con, gth_sa_ref_seq_num(sa));
}
switch (gth_sa_alphatype(sa)) {
case DNA_ALPHA:
/* compute the two alignment lines */
cols = gthfillthetwoalignmentlines(first_line,
second_line,
gen_seq_orig +
genomicstartcutoff,
gth_sa_gen_dp_length(sa) -
genomictotalcutoff,
ref_seq_orig +
referencestartcutoff,
gth_sa_ref_total_length(sa) -
referencetotalcutoff,
gth_sa_get_editoperations(sa),
gth_sa_get_editoperations_length(sa),
0, /* linewidth not important here */
0, /* no short introns here */
NULL,/* therefore no shortintroninfo */
gth_sa_indelcount(sa));
*third_line = NULL;
break;
case PROTEIN_ALPHA:
/* compute the three alignment lines */
cols = gthfillthethreealignmentlines(first_line,
second_line,
third_line,
gth_sa_get_editoperations(sa),
gth_sa_get_editoperations_length(sa),
gth_sa_indelcount(sa),
gen_seq_orig +
genomicstartcutoff,
gth_sa_gen_dp_length(sa) -
genomictotalcutoff,
ref_seq_orig +
referencestartcutoff,
gth_sa_ref_total_length(sa) -
referencetotalcutoff,
translationtable);
break;
default: gt_assert(0);
}
return cols;
}
void gth_sa_echo_alignment(const GthSA *sa, GtUword showintronmaxlen,
GtUword translationtable,
bool wildcardimplosion, GthInput *input,
GtFile *outfp)
{
GtUword genomicstartcutoff, genomicendcutoff, genomictotalcutoff,
referencestartcutoff, referenceendcutoff, referencetotalcutoff;
bool reverse_subject_pos = false;
const unsigned char *gen_seq_orig, *ref_seq_orig;
GthSeqCon *ref_seq_con;
GtAlphabet *ref_alphabet;
gt_assert(sa && input);
/* only for cosmetic reasons */
genomicstartcutoff = gth_sa_genomiccutoff_start(sa);
genomicendcutoff = gth_sa_genomiccutoff_end(sa);
genomictotalcutoff = genomicstartcutoff + genomicendcutoff;
referencestartcutoff = gth_sa_referencecutoff_start(sa);
referenceendcutoff = gth_sa_referencecutoff_end(sa);
referencetotalcutoff = referencestartcutoff + referenceendcutoff;
/* make sure that the correct files are loaded */
gth_input_load_reference_file(input, gth_sa_ref_file_num(sa), false);
ref_seq_con = gth_input_current_ref_seq_con(input);
ref_alphabet = gth_input_current_ref_alphabet(input);
/* If the reverse complement of the genomic DNA is considered, this
opition is needed for correct output of the genomic sequence positions
by the function showalignmentgeneric() */
if (!gth_sa_gen_strand_forward(sa))
reverse_subject_pos = true;
/* get genomic sequence */
gen_seq_orig =
gth_input_original_genomic_sequence(input, sa->gen_file_num,
sa->gen_strand_forward)
+ gth_sa_gen_dp_start(sa);
/* get reference sequence */
if (gth_sa_ref_strand_forward(sa)) {
ref_seq_orig =
gth_seq_con_get_orig_seq(ref_seq_con, gth_sa_ref_seq_num(sa));
}
else {
ref_seq_orig =
gth_seq_con_get_orig_seq_rc(ref_seq_con, gth_sa_ref_seq_num(sa));
}
switch (gth_sa_alphatype(sa)) {
case DNA_ALPHA:
gthshowalignmentdna(outfp,ALIGNMENTLINEWIDTH,
gth_sa_get_editoperations(sa),
gth_sa_get_editoperations_length(sa),
gth_sa_indelcount(sa),
gen_seq_orig + genomicstartcutoff,
gth_sa_gen_dp_length(sa) - genomictotalcutoff,
ref_seq_orig + referencestartcutoff,
gth_sa_ref_total_length(sa) -
referencetotalcutoff,
gth_sa_gen_dp_start(sa) + genomicstartcutoff -
gth_sa_gen_offset(sa), referencestartcutoff,
gth_sa_gen_total_length(sa), showintronmaxlen,
ref_alphabet, reverse_subject_pos,
wildcardimplosion);
break;
case PROTEIN_ALPHA:
gthshowalignmentprotein(outfp, ALIGNMENTLINEWIDTH,
gth_sa_get_editoperations(sa),
gth_sa_get_editoperations_length(sa),
gth_sa_indelcount(sa),
gen_seq_orig + genomicstartcutoff,
gth_sa_gen_dp_length(sa) - genomictotalcutoff,
ref_seq_orig + referencestartcutoff,
gth_sa_ref_total_length(sa) -
referencetotalcutoff,
gth_sa_gen_dp_start(sa) + genomicstartcutoff -
gth_sa_gen_offset(sa), referencestartcutoff,
gth_sa_gen_total_length(sa), showintronmaxlen,
ref_alphabet, translationtable,
gth_input_score_matrix(input),
gth_input_score_matrix_alpha(input),
reverse_subject_pos, wildcardimplosion);
break;
default: gt_assert(0);
}
}
