int splicing_gff_noexons_one(const splicing_gff_t *gff, size_t gene,
splicing_vector_int_t *noexons) {
size_t nogenes, idx1, idx2, noiso, pos, il;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
if (gene >= nogenes) {
SPLICING_ERROR("Invalid gene id", SPLICING_EINVAL);
}
idx1=VECTOR(gff->genes)[gene];
idx2= gene+1 == nogenes ? gff->n : VECTOR(gff->genes)[gene+1];
for (noiso=0; idx1 < idx2; idx1++) {
if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_MRNA) { noiso += 1; }
}
SPLICING_CHECK(splicing_vector_int_resize(noexons, noiso));
idx1=VECTOR(gff->genes)[gene];
idx2= gene+1 == nogenes ? gff->n : VECTOR(gff->genes)[gene+1];
for (; idx1 < idx2 && VECTOR(gff->type)[idx1] != SPLICING_TYPE_MRNA;
idx1++) ;
idx1++;
for (pos=0, il=0; idx1 < idx2; idx1++) {
if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_MRNA) {
VECTOR(*noexons)[pos++]=il;
il=0;
} else if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_EXON) { il++; }
}
VECTOR(*noexons)[pos++]=il;
return 0;
}
int splicing_gff_exon_start_end(const splicing_gff_t *gff,
splicing_vector_int_t *start,
splicing_vector_int_t *end,
splicing_vector_int_t *idx,
int gene) {
size_t noiso;
int i=0, p=0, n=splicing_gff_size(gff);
int pos;
size_t nogenes;
splicing_vector_int_t tmp, tmp2;
SPLICING_CHECK(splicing_vector_int_init(&tmp, 10));
SPLICING_FINALLY(splicing_vector_int_destroy, &tmp);
SPLICING_CHECK(splicing_vector_int_init(&tmp2, 10));
SPLICING_FINALLY(splicing_vector_int_destroy, &tmp2);
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
if (gene < 0 || gene >= nogenes) {
SPLICING_ERROR("Invalid gene id", SPLICING_EINVAL);
}
pos=VECTOR(gff->genes)[gene]+1;
SPLICING_CHECK(splicing_gff_noiso_one(gff, gene, &noiso));
splicing_vector_int_clear(start);
splicing_vector_int_clear(end);
SPLICING_CHECK(splicing_vector_int_resize(idx, noiso+1));
while (pos < n) {
if (VECTOR(gff->type)[pos] == SPLICING_TYPE_EXON) {
int s=VECTOR(gff->start)[pos];
int e=VECTOR(gff->end)[pos];
SPLICING_CHECK(splicing_vector_int_push_back(start, s)); p++;
SPLICING_CHECK(splicing_vector_int_push_back(end, e));
} else if (VECTOR(gff->type)[pos] == SPLICING_TYPE_MRNA) {
VECTOR(*idx)[i] = p;
if (i!=0) {
SPLICING_CHECK(splicing_i_gff_exon_start_end_sort(start, end, idx,
i-1, &tmp, &tmp2));
}
i++;
} else if (VECTOR(gff->type)[pos] == SPLICING_TYPE_GENE) {
break;
}
pos++;
}
VECTOR(*idx)[i] = p;
SPLICING_CHECK(splicing_i_gff_exon_start_end_sort(start, end, idx, i-1,
&tmp, &tmp2));
splicing_vector_int_destroy(&tmp2);
splicing_vector_int_destroy(&tmp);
SPLICING_FINALLY_CLEAN(1);
return 0;
}
int splicing_i_gff_constitutive_exons_all(const splicing_gff_t *gff,
splicing_exonset_t *exons,
int min_length) {
size_t g, nogenes;
splicing_vector_int_t events;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
SPLICING_CHECK(splicing_exonset_init(exons, 0));
SPLICING_FINALLY(splicing_exonset_destroy, exons);
SPLICING_CHECK(splicing_vector_int_init(&events, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &events);
for (g=0; g<nogenes; g++) {
const char *seqid=
splicing_strvector_get(&gff->seqids, VECTOR(gff->seqid)[g]);
int noex, idx, noEvents, i;
size_t noiso;
int start=VECTOR(gff->genes)[g];
int end= g+1 < nogenes ? VECTOR(gff->genes)[g+1] : gff->n;
splicing_gff_noiso_one(gff, g, &noiso);
/* Collect and sort all events */
splicing_vector_int_clear(&events);
for (idx=start+1; idx<end; idx++) {
if (VECTOR(gff->type)[idx] == SPLICING_TYPE_EXON) {
SPLICING_CHECK(splicing_vector_int_push_back2
(&events, VECTOR(gff->start)[idx],
-VECTOR(gff->end)[idx]));
}
}
noEvents=splicing_vector_int_size(&events);
splicing_qsort(VECTOR(events), noEvents, sizeof(int),
splicing_i_const_cmp);
/* Now go over the sorted events and extract the constitutive exons */
for (noex=0, i=0; i<noEvents; i++) {
int ev=VECTOR(events)[i];
if (ev > 0) { noex++; }
if (ev < 0) {
int prev=VECTOR(events)[i-1];
if (noex == noiso && (-ev)-prev+1 >= min_length) {
/* constitutive exon */
SPLICING_CHECK(splicing_exonset_append(exons, seqid, prev, -ev));
}
noex--;
}
}
}
splicing_vector_int_destroy(&events);
SPLICING_FINALLY_CLEAN(2); /* + exons */
return 0;
}
int splicing_i_gff_constitutive_exons_full(const splicing_gff_t *gff,
splicing_exonset_t *exons,
int min_length) {
size_t g, nogenes;
splicing_vector_int_t events;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
SPLICING_CHECK(splicing_exonset_init(exons, 1));
SPLICING_FINALLY(splicing_exonset_destroy, exons);
SPLICING_CHECK(splicing_vector_int_init(&events, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &events);
for (g=0; g<nogenes; g++) {
const char *seqid=
splicing_strvector_get(&gff->seqids, VECTOR(gff->seqid)[g]);
int i, idx, noExons, noSame;
size_t noiso;
int start=VECTOR(gff->genes)[g];
int end= g+1 < nogenes ? VECTOR(gff->genes)[g+1] : gff->n;
splicing_gff_noiso_one(gff, g, &noiso);
/* Collect and sort all events */
splicing_vector_int_clear(&events);
for (idx=start+1; idx<end; idx++) {
if (VECTOR(gff->type)[idx] == SPLICING_TYPE_EXON) {
SPLICING_CHECK(splicing_vector_int_push_back2
(&events, VECTOR(gff->start)[idx],
VECTOR(gff->end)[idx]));
}
}
noExons=splicing_vector_int_size(&events)/2;
splicing_qsort(VECTOR(events), noExons, sizeof(int)*2,
splicing_i_const_cmp2);
/* Now go over them and check how many times each exon appears */
for (noSame=1, i=2; i<noExons*2; i+=2) {
int start=VECTOR(events)[i];
int end=VECTOR(events)[i+1];
if (start == VECTOR(events)[i-2] && VECTOR(events)[i-1] == end) {
noSame++;
} else {
noSame=1;
}
if (noSame == noiso) {
SPLICING_CHECK(splicing_exonset_append(exons, seqid, start, end));
}
}
}
splicing_vector_int_destroy(&events);
SPLICING_FINALLY_CLEAN(2); /* + exons */
return 0;
}
int splicing_gff_fprint(const splicing_gff_t *gff,
FILE *outfile) {
size_t i, n;
SPLICING_CHECK(splicing_gff_nogenes(gff, &n));
for (i=0; i<n; i++) {
SPLICING_CHECK(splicing_gff_fprint_gene(gff, outfile, i));
}
return 0;
}
static PyObject* pysplicing_gff_nogenes(PyObject *self, PyObject *args) {
PyObject *gff;
size_t nogenes;
int nogenes2;
splicing_gff_t *mygff;
if (!PyArg_ParseTuple(args, "O", &gff)) { return NULL; }
mygff=PyCObject_AsVoidPtr(gff);
SPLICING_PYCHECK(splicing_gff_nogenes(mygff, &nogenes));
nogenes2=nogenes; /* in case int and size_t are different */
return Py_BuildValue("i", nogenes2);
}
int splicing_i_gff_noiso_one(const splicing_gff_t *gff, size_t gene,
size_t *noiso, splicing_vector_int_t *isolen) {
size_t nogenes, idx1, idx2;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
if (gene < 0 || gene >= nogenes) {
SPLICING_ERROR("Invalid gene id", SPLICING_EINVAL);
}
idx1=VECTOR(gff->genes)[gene];
idx2= gene+1 == nogenes ? gff->n : VECTOR(gff->genes)[gene+1];
*noiso = 0;
for ( ; idx1 < idx2; idx1++) {
if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_MRNA) { *noiso += 1; }
}
if (isolen) {
size_t il=0, pos=0;
SPLICING_CHECK(splicing_vector_int_resize(isolen, *noiso));
idx1=VECTOR(gff->genes)[gene];
idx2= gene+1 == nogenes ? gff->n : VECTOR(gff->genes)[gene+1];
for (; idx1 < idx2 && VECTOR(gff->type)[idx1] != SPLICING_TYPE_MRNA;
idx1++) ;
idx1++;
for (; idx1 < idx2; idx1++) {
if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_MRNA) {
VECTOR(*isolen)[pos++]=il;
il = 0;
} else if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_EXON) {
il += VECTOR(gff->end)[idx1] - VECTOR(gff->start)[idx1] + 1;
}
}
VECTOR(*isolen)[pos++]=il;
}
return 0;
}
int splicing_gff_noiso(const splicing_gff_t *gff,
splicing_vector_int_t *noiso) {
size_t nogenes, idx1, idx2, pos=0;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
idx1=VECTOR(gff->genes)[0];
idx2=gff->n;
SPLICING_CHECK(splicing_vector_int_resize(noiso, nogenes));
splicing_vector_int_null(noiso);
if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_GENE) { idx1++; }
for (; idx1 < gff->n; idx1++) {
if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_MRNA) {
VECTOR(*noiso)[pos] += 1;
} else if (VECTOR(gff->type)[idx1] == SPLICING_TYPE_GENE) {
pos++;
}
}
return 0;
}
int splicing_gff_fprint_gene(const splicing_gff_t *gff,
FILE *outfile, int gene) {
size_t nogenes, noiso;
int i, j;
splicing_vector_int_t start, end, idx;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
if (gene < 0 || gene >= nogenes) {
SPLICING_ERROR("Invalid gene ID", SPLICING_EINVAL);
}
SPLICING_CHECK(splicing_vector_int_init(&start, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &start);
SPLICING_CHECK(splicing_vector_int_init(&end, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &end);
SPLICING_CHECK(splicing_vector_int_init(&idx, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &idx);
SPLICING_CHECK(splicing_gff_exon_start_end(gff, &start, &end, &idx, gene));
noiso = splicing_vector_int_size(&idx)-1;
fprintf(outfile, "===\nGene with %i isoforms:\n", (int) noiso);
for (i=0; i<noiso; i++) {
fprintf(outfile, " Isoform %i:\n", i);
for (j=VECTOR(idx)[i]; j<VECTOR(idx)[i+1]; j++) {
fprintf(outfile, " %i-%i\n", VECTOR(start)[j], VECTOR(end)[j]);
}
}
splicing_vector_int_destroy(&idx);
splicing_vector_int_destroy(&end);
splicing_vector_int_destroy(&start);
SPLICING_FINALLY_CLEAN(3);
return 0;
}
int splicing_simulate_reads(const splicing_gff_t *gff, int gene,
const splicing_vector_t *expression,
int noreads, int readLength,
splicing_vector_int_t *isoform,
splicing_vector_int_t *position,
splicing_strvector_t *cigar,
splicing_vector_t *sample_prob) {
size_t i, p, noiso, goodiso=0, nogenes;
splicing_vector_int_t effisolen;
splicing_vector_t sampleprob;
double rand, sumpsi=0.0;
splicing_vector_int_t exstart, exend, exidx;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
if (gene < 0 || gene >= nogenes) {
SPLICING_ERROR("Invalid gene id", SPLICING_EINVAL);
}
/* TODO: more error checks */
SPLICING_CHECK(splicing_gff_noiso_one(gff, gene, &noiso));
SPLICING_CHECK(splicing_vector_int_init(&effisolen, noiso));
SPLICING_FINALLY(splicing_vector_int_destroy, &effisolen);
SPLICING_CHECK(splicing_vector_init(&sampleprob, noiso));
SPLICING_FINALLY(splicing_vector_destroy, &sampleprob);
SPLICING_CHECK(splicing_vector_int_resize(isoform, noreads));
SPLICING_CHECK(splicing_gff_isolength_one(gff, gene, &effisolen));
for (i=0; i<noiso; i++) {
int l=VECTOR(effisolen)[i]-readLength+1;
VECTOR(effisolen)[i] = l > 0 ? l : 0;
VECTOR(sampleprob)[i] = VECTOR(*expression)[i] * VECTOR(effisolen)[i];
if (VECTOR(sampleprob)[i] != 0) { goodiso++; }
sumpsi += VECTOR(sampleprob)[i];
}
if (goodiso==0) {
SPLICING_ERROR("No isoform is possible", SPLICING_FAILURE);
}
if (sample_prob) {
SPLICING_CHECK(splicing_vector_update(sample_prob, &sampleprob));
}
for (i=1; i<noiso; i++) {
VECTOR(sampleprob)[i] += VECTOR(sampleprob)[i-1];
}
for (i=0; i<noreads; i++) {
int w;
if (noiso==1) {
w=0;
} else if (noiso==2) {
rand = RNG_UNIF01() * sumpsi;
w = (rand < VECTOR(sampleprob)[0]) ? 0 : 1;
} else {
rand = RNG_UNIF01() * sumpsi;
for (w=0; rand > VECTOR(sampleprob)[w]; w++) ;
}
VECTOR(*isoform)[i]=w;
}
splicing_vector_destroy(&sampleprob);
SPLICING_FINALLY_CLEAN(1);
/* OK, we have the isoforms, now we need the read positions,
these are uniformly sampled from the individual isoforms. */
SPLICING_CHECK(splicing_vector_int_resize(position, noreads));
SPLICING_CHECK(splicing_vector_int_init(&exstart, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &exstart);
SPLICING_CHECK(splicing_vector_int_init(&exend, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &exend);
SPLICING_CHECK(splicing_vector_int_init(&exidx, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &exidx);
SPLICING_CHECK(splicing_gff_exon_start_end(gff, &exstart, &exend, &exidx,
gene));
/* Positions in isoform coordinates first */
for (i=0; i<noreads; i++) {
int iso=VECTOR(*isoform)[i];
int len=VECTOR(effisolen)[iso];
VECTOR(*position)[i]=RNG_INTEGER(1, len);
}
/* Translate isoform coordinates to genomic coordintes */
/* TODO: some of this is already calculated */
SPLICING_CHECK(splicing_iso_to_genomic(gff, gene, isoform, /*converter=*/ 0,
position));
/* CIGAR strings */
splicing_strvector_clear(cigar);
SPLICING_CHECK(splicing_strvector_reserve(cigar, noreads));
for (i=0; i<noreads; i++) {
char tmp[1000], *tmp2=tmp;
int iso=VECTOR(*isoform)[i];
size_t rs=VECTOR(*position)[i];
int ex=0;
int rl=readLength;
for (ex=VECTOR(exidx)[iso]; VECTOR(exend)[ex] < rs; ex++) ;
while (VECTOR(exend)[ex] < rs+rl-1) {
tmp2 += snprintf(tmp2, sizeof(tmp)/sizeof(char)-(tmp2-tmp)-1, "%iM%iN",
(int) (VECTOR(exend)[ex]-rs+1),
(int) (VECTOR(exstart)[ex+1]-VECTOR(exend)[ex]-1));
if (tmp2 >= tmp + sizeof(tmp)/sizeof(char)) {
SPLICING_ERROR("CIGAR string too long", SPLICING_EINVAL);
}
rl -= (VECTOR(exend)[ex] - rs + 1);
rs = VECTOR(exstart)[ex+1];
ex++;
}
tmp2 += snprintf(tmp2, sizeof(tmp)/sizeof(char)-(tmp2-tmp)-1, "%iM", rl);
if (tmp2 >= tmp + sizeof(tmp)/sizeof(char)) {
SPLICING_ERROR("CIGAR string too long", SPLICING_EINVAL); }
SPLICING_CHECK(splicing_strvector_append(cigar, tmp));
}
splicing_vector_int_destroy(&exidx);
splicing_vector_int_destroy(&exend);
splicing_vector_int_destroy(&exstart);
splicing_vector_int_destroy(&effisolen);
SPLICING_FINALLY_CLEAN(4);
return 0;
}
int splicing_simulate_paired_reads(const splicing_gff_t *gff, int gene,
const splicing_vector_t *expression,
int noreads, int readLength,
const splicing_vector_t *fragmentProb,
int fragmentStart, double normalMean,
double normalVar, double numDevs,
splicing_vector_int_t *isoform,
splicing_vector_int_t *position,
splicing_strvector_t *cigar,
splicing_vector_t *sampleprob) {
size_t i, j, noiso, il, nogenes;
splicing_vector_t *mysampleprob=sampleprob, vsampleprob;
splicing_vector_t px, cpx;
double sumpx, sumpsi=0.0;
splicing_vector_int_t isolen;
int goodiso=0;
splicing_vector_int_t exstart, exend, exidx;
splicing_vector_t *myfragmentProb=(splicing_vector_t*) fragmentProb,
vfragmentProb;
int fs, fl;
SPLICING_CHECK(splicing_gff_nogenes(gff, &nogenes));
if (gene < 0 || gene >= nogenes) {
SPLICING_ERROR("Invalid gene id", SPLICING_EINVAL);
}
/* TODO: more error checks */
if (!fragmentProb) {
myfragmentProb=&vfragmentProb;
SPLICING_CHECK(splicing_vector_init(&vfragmentProb, 0));
SPLICING_FINALLY(splicing_vector_destroy, &vfragmentProb);
SPLICING_CHECK(splicing_normal_fragment(normalMean, normalVar, numDevs,
readLength, myfragmentProb,
&fragmentStart));
splicing_vector_scale(myfragmentProb,
1.0/splicing_vector_sum(myfragmentProb));
}
il=splicing_vector_size(myfragmentProb);
fs=fragmentStart;
fl=fragmentStart+il-1;
SPLICING_CHECK(splicing_gff_noiso_one(gff, gene, &noiso));
if ( fabs(splicing_vector_sum(myfragmentProb) - 1.0) > 1e-10 ) {
SPLICING_ERROR("Fragment length distribution does not sum up to 1",
SPLICING_EINVAL);
}
SPLICING_CHECK(splicing_vector_int_init(&isolen, noiso));
SPLICING_FINALLY(splicing_vector_int_destroy, &isolen);
SPLICING_CHECK(splicing_gff_isolength_one(gff, gene, &isolen));
SPLICING_CHECK(splicing_vector_copy(&px, myfragmentProb));
SPLICING_FINALLY(splicing_vector_destroy, &px);
SPLICING_CHECK(splicing_vector_init(&cpx, il));
SPLICING_FINALLY(splicing_vector_destroy, &cpx);
if (!sampleprob) {
mysampleprob=&vsampleprob;
SPLICING_CHECK(splicing_vector_init(mysampleprob, noiso));
SPLICING_FINALLY(splicing_vector_destroy, mysampleprob);
} else {
SPLICING_CHECK(splicing_vector_resize(mysampleprob, noiso));
}
for (sumpx=VECTOR(px)[0], i=1; i<il; i++) {
VECTOR(px)[i] += VECTOR(px)[i-1];
sumpx += VECTOR(px)[i];
}
VECTOR(cpx)[0] = VECTOR(px)[0];
for (i=1; i<il; i++) {
VECTOR(cpx)[i] = VECTOR(cpx)[i-1] + VECTOR(px)[i];
}
for (i=0; i<noiso; i++) {
int ilen=VECTOR(isolen)[i];
int r1= ilen >= fl ? ilen - fl + 1 : 0;
int r2= ilen >= fs ? (ilen >= fl ? fl - fs : ilen - fs + 1) : 0;
/* int r3= fs - 1; */
double sp=0.0;
if (r1 > 0) { sp += r1; }
if (r2 > 0) { sp += VECTOR(cpx)[r2-1]; }
VECTOR(*mysampleprob)[i] = sp * VECTOR(*expression)[i];
if (VECTOR(*mysampleprob)[i] != 0) { goodiso += 1; }
sumpsi += VECTOR(*mysampleprob)[i];
}
if (goodiso == 0) {
SPLICING_ERROR("No isoform is possible", SPLICING_FAILURE);
}
for (i=1; i<noiso; i++) {
VECTOR(*mysampleprob)[i] += VECTOR(*mysampleprob)[i-1];
}
SPLICING_CHECK(splicing_vector_int_resize(isoform, noreads*2));
for (i=0; i<2*noreads; i+=2) {
int w;
double rand;
if (noiso==1) {
w=0;
} else if (noiso==2) {
rand = RNG_UNIF01() * sumpsi;
w = (rand < VECTOR(*mysampleprob)[0]) ? 0 : 1;
} else {
rand = RNG_UNIF01() * sumpsi;
for (w=0; rand > VECTOR(*mysampleprob)[w]; w++) ;
}
VECTOR(*isoform)[i]=VECTOR(*isoform)[i+1]=w;
}
if (!sampleprob) {
splicing_vector_destroy(mysampleprob);
SPLICING_FINALLY_CLEAN(1);
} else {
for (i=noiso-1; i>0; i--) {
VECTOR(*mysampleprob)[i] -= VECTOR(*mysampleprob)[i-1];
}
}
/* We have the isoforms, now get the read positions. */
SPLICING_CHECK(splicing_vector_int_resize(position, noreads*2));
SPLICING_CHECK(splicing_vector_int_init(&exstart, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &exstart);
SPLICING_CHECK(splicing_vector_int_init(&exend, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &exend);
SPLICING_CHECK(splicing_vector_int_init(&exidx, 0));
SPLICING_FINALLY(splicing_vector_int_destroy, &exidx);
SPLICING_CHECK(splicing_gff_exon_start_end(gff, &exstart, &exend, &exidx,
gene));
/* Positions in isoform coordinates first.
These are sampled based on the fragment length distribution. */
for (i=0, j=0; i<noreads; i++) {
int iso=VECTOR(*isoform)[2*i];
int ilen=VECTOR(isolen)[iso];
int r1= ilen >= fl ? ilen - fl + 1 : 0;
int r2= ilen >= fs ? (ilen >= fl ? fl - fs : ilen - fs + 1) : 0;
/* int r3= fs - 1; */
int pos, fragment;
double sp=0.0;
if (r1 > 0) { sp += r1; }
if (r2 > 0) { sp += VECTOR(cpx)[r2-1]; }
double rand=RNG_UNIF(0, sp);
if (rand < r1) {
pos = ceil(rand);
} else {
int w;
rand -= r1;
for (w=0; VECTOR(cpx)[w] < rand; w++) ;
pos = r1 + r2 - w;
}
if (pos <= r1) {
rand=RNG_UNIF(0, 1.0);
} else {
rand=RNG_UNIF(0, VECTOR(px)[r1+r2-pos]);
}
for (fragment=0; VECTOR(px)[fragment] < rand; fragment++) ;
fragment += fragmentStart;
VECTOR(*position)[j++] = pos;
VECTOR(*position)[j++] = pos+fragment-readLength;
}
/* Translate positions to genomic coordinates */
/* TODO: some of this is already calculated */
SPLICING_CHECK(splicing_iso_to_genomic(gff, gene, isoform, /*converter=*/ 0,
position));
/* CIGAR strings */
splicing_strvector_clear(cigar);
SPLICING_CHECK(splicing_strvector_reserve(cigar, 2*noreads));
for (j=0; j<2*noreads; j++) {
char tmp[1000], *tmp2=tmp;
int iso=VECTOR(*isoform)[j];
size_t rs=VECTOR(*position)[j];
int ex=0;
int rl=readLength;
for (ex=VECTOR(exidx)[iso]; VECTOR(exend)[ex] < rs; ex++) ;
while (rs + rl - 1 > VECTOR(exend)[ex]) {
tmp2 += snprintf(tmp2, sizeof(tmp)/sizeof(char)-(tmp2-tmp)-1, "%iM%iN",
(int) (VECTOR(exend)[ex]-rs+1),
(int) (VECTOR(exstart)[ex+1]-VECTOR(exend)[ex]-1));
if (tmp2 >= tmp + sizeof(tmp)/sizeof(char)) {
SPLICING_ERROR("CIGAR string too long", SPLICING_EINVAL);
}
rl -= (VECTOR(exend)[ex] - rs + 1);
rs = VECTOR(exstart)[ex+1];
ex++;
}
tmp2 += snprintf(tmp2, sizeof(tmp)/sizeof(char)-(tmp2-tmp)-1, "%iM", rl);
if (tmp2 >= tmp + sizeof(tmp)/sizeof(char)) {
SPLICING_ERROR("CIGAR string too long", SPLICING_EINVAL);
}
SPLICING_CHECK(splicing_strvector_append(cigar, tmp));
}
splicing_vector_int_destroy(&exidx);
splicing_vector_int_destroy(&exend);
splicing_vector_int_destroy(&exstart);
splicing_vector_destroy(&cpx);
splicing_vector_destroy(&px);
splicing_vector_int_destroy(&isolen);
SPLICING_FINALLY_CLEAN(6);
if (!fragmentProb) {
splicing_vector_destroy(myfragmentProb);
SPLICING_FINALLY_CLEAN(1);
}
return 0;
}
