inline void PROBerReadModel::update(InMemAlignG* ag_in_mem, InMemAlign* aligns, AlignmentGroup& ag, double noise_frac) {
SEQstring seq;
QUALstring qual;
CIGARstring cigar;
const RefSeq* refseq = NULL;
assert(ag.getSEQ(seq));
if (model_type & 1) assert(ag.getQUAL(qual));
// update noise prob
npro->update(seq, noise_frac);
// update alignment probs
for (int i = 0; i < ag_in_mem->size; ++i) if (aligns[i].frac > 0.0) {
refseq = refs->getRef(aligns[i].tid);
assert(ag.getAlignment(i)->getCIGAR(cigar));
seqmodel->update(aligns[i].frac, '+', aligns[i].pos, refseq, &cigar, &seq, ((model_type & 1) ? &qual : NULL));
}
if (model_type >= 2) {
// paired-end reads
assert(ag.getSEQ(seq, 2));
if (model_type & 1) assert(ag.getQUAL(qual, 2));
// update noise prob
npro->update(seq, noise_frac);
// update alignment probs
for (int i = 0; i < ag_in_mem->size; ++i) if (aligns[i].frac > 0.0) {
refseq = refs->getRef(aligns[i].tid);
assert(ag.getAlignment(i)->getCIGAR(cigar, 2));
assert(aligns[i].fragment_length > 0);
seqmodel->update(aligns[i].frac, '-', refseq->getLen() - aligns[i].pos - aligns[i].fragment_length, refseq, &cigar, &seq, ((model_type & 1) ? &qual : NULL));
}
}
}
void calcExpectedEffectiveLengths(ModelType& model) {
int lb, ub, span;
double *pdf = NULL, *cdf = NULL, *clen = NULL; // clen[i] = sigma_{j=1}^{i}pdf[i]*(lb+i)
model.getGLD().copyTo(pdf, cdf, lb, ub, span);
clen = new double[span + 1];
clen[0] = 0.0;
for (int i = 1; i <= span; i++) {
clen[i] = clen[i - 1] + pdf[i] * (lb + i);
}
eel.clear();
eel.resize(M + 1, 0.0);
for (int i = 1; i <= M; i++) {
int totLen = refs.getRef(i).getTotLen();
int fullLen = refs.getRef(i).getFullLen();
int pos1 = max(min(totLen - fullLen + 1, ub) - lb, 0);
int pos2 = max(min(totLen, ub) - lb, 0);
if (pos2 == 0) { eel[i] = 0.0; continue; }
eel[i] = fullLen * cdf[pos1] + ((cdf[pos2] - cdf[pos1]) * (totLen + 1) - (clen[pos2] - clen[pos1]));
assert(eel[i] >= 0);
if (eel[i] < MINEEL) { eel[i] = 0.0; }
}
delete[] pdf;
delete[] cdf;
delete[] clen;
}
inline void PROBerReadModel::setConProbs(InMemAlignG* ag_in_mem, InMemAlign* aligns, AlignmentGroup& ag) {
int seqlen;
SEQstring seq; // seq, qual and cigar must be in each function since we have multiple threads!
QUALstring qual;
CIGARstring cigar;
const RefSeq* refseq = NULL;
// Get read sequences and quality scores
assert(ag.getSEQ(seq));
seqlen = read_length < 0 ? ag.getSeqLength() : read_length;
if (model_type & 1) assert(ag.getQUAL(qual));
// set noise probability
ag_in_mem->noise_conprb = mld1->getProb(seqlen) * npro->getProb(seq);
// set alignment probabilities
for (int i = 0; i < ag_in_mem->size; ++i) if (aligns[i].conprb != -1.0) {
refseq = refs->getRef(aligns[i].tid);
assert(ag.getAlignment(i)->getCIGAR(cigar));
aligns[i].conprb = (aligns[i].fragment_length > 0 ? mld1->getProb(seqlen, aligns[i].fragment_length) : mld1->getProb(seqlen)) * \
seqmodel->getProb('+', aligns[i].pos, refseq, &cigar, &seq, ((model_type & 1) ? &qual : NULL));
}
if (model_type >= 2) {
// paired-end reads
assert(ag.getSEQ(seq, 2));
seqlen = read_length < 0 ? ag.getSeqLength(2) : read_length;
if (model_type & 1) assert(ag.getQUAL(qual, 2));
ag_in_mem->noise_conprb *= mld2->getProb(seqlen) * npro->getProb(seq);
for (int i = 0; i < ag_in_mem->size; ++i) if (aligns[i].conprb != -1.0) {
refseq = refs->getRef(aligns[i].tid);
assert(ag.getAlignment(i)->getCIGAR(cigar, 2));
assert(aligns[i].fragment_length > 0);
aligns[i].conprb *= mld2->getProb(seqlen, aligns[i].fragment_length) * \
seqmodel->getProb('-', refseq->getLen() - aligns[i].pos - aligns[i].fragment_length, refseq, &cigar, &seq, ((model_type & 1) ? &qual : NULL));
}
}
}
inline void PROBerReadModel::simulate(READ_INT_TYPE rid, int tid, int pos, int fragment_length, std::ofstream* out1, std::ofstream* out2) {
int m2pos;
int mateL1, mateL2;
std::string qual1, qual2, cigar1, cigar2, readseq1, readseq2;
// Simulate reads
if (tid == 0) {
cigar1 = cigar2 = "*";
mateL1 = mld1->simulate(sampler);
if (model_type & 1) qd->simulate(sampler, mateL1, qual1);
npro->simulate(sampler, mateL1, readseq1);
if (model_type >= 2) {
mateL2 = mld2->simulate(sampler);
if (model_type & 1) qd->simulate(sampler, mateL2, qual2);
npro->simulate(sampler, mateL2, readseq2);
}
}
else {
const RefSeq* ref = refs->getRef(tid);
mateL1 = mld1->simulate(sampler, fragment_length);
if (model_type & 1) qd->simulate(sampler, mateL1, qual1);
seqmodel->simulate(sampler, mateL1, '+', pos, ref, qual1, cigar1, readseq1);
if (model_type >= 2) {
mateL2 = mld2->simulate(sampler, fragment_length);
if (model_type & 1) qd->simulate(sampler, mateL2, qual2);
m2pos = ref->getLen() - pos - fragment_length;
seqmodel->simulate(sampler, mateL2, '-', m2pos, ref, qual2, cigar2, readseq2);
}
}
// Output reads
(*out1)<< ((model_type & 1) ? '@' : '>') << rid<< '_'<< tid<< '_'<< pos<< '_'<< fragment_length<< '_'<< cigar1<< (model_type >= 2 ? "/1" : "") << std::endl;
(*out1)<< readseq1<< std::endl;
if (model_type & 1) (*out1)<< '+'<< std::endl<< qual1<< std::endl;
if (model_type >= 2) {
(*out2)<< ((model_type & 1) ? '@' : '>') << rid<< '_'<< tid<< '_'<< pos<< '_'<< fragment_length<< '_'<< cigar2<< "/2"<< std::endl;
(*out2)<< readseq2<< std::endl;
if (model_type & 1) (*out2)<< '+'<< std::endl<< qual2<< std::endl;
}
}
void writeToDisk(char* refName) {
ofstream fout;
sprintf(tiF, "%s.ti", refName);
transcripts.writeTo(tiF);
if (verbose) { printf("Transcript Information File is generated!\n"); }
sprintf(refFastaF, "%s.transcripts.fa", refName);
refs.writeTo(refFastaF);
sprintf(transListF, "%s.translist", refName);
refs.writeTransListTo(transListF);
sprintf(chromListF, "%s.chrlist", refName);
fout.open(chromListF);
for (int i = 0; i < (int)chrvec.size(); ++i)
fout<< chrvec[i].name<< '\t'<< chrvec[i].len<< endl;
fout.close();
if (verbose) { printf("Chromosome List File is generated!\n"); }
string cur_gene_id, cur_transcript_id, name;
vector<int> gi, gt, ta;
cur_gene_id = ""; gi.clear();
if (mappingType == 2) { cur_transcript_id = ""; gt.clear(); ta.clear(); }
for (int i = 1; i <= M; ++i) {
const Transcript& transcript = transcripts.getTranscriptAt(i);
if (cur_gene_id != transcript.getGeneID()) {
gi.push_back(i);
if (mappingType == 2) gt.push_back((int)ta.size());
cur_gene_id = transcript.getGeneID();
}
if ((mappingType == 2) && (cur_transcript_id != transcript.getTranscriptID())) {
ta.push_back(i);
cur_transcript_id = transcript.getTranscriptID();
}
}
gi.push_back(M + 1);
if (mappingType == 2) { gt.push_back((int)ta.size()); ta.push_back(M + 1); }
sprintf(groupF, "%s.grp", refName);
fout.open(groupF);
for (int i = 0; i < (int)gi.size(); ++i) fout<< gi[i]<< endl;
fout.close();
if (verbose) { printf("Group File is generated!\n"); }
if (mappingType == 2) {
sprintf(gtF, "%s.gt", refName);
fout.open(gtF);
for (int i = 0; i < (int)gt.size(); ++i) fout<< gt[i]<< endl;
fout.close();
sprintf(taF, "%s.ta", refName);
fout.open(taF);
for (int i = 0; i < (int)ta.size(); ++i) fout<< ta[i]<< endl;
fout.close();
if (verbose) { printf("Allele-specific group files are generated!\n"); }
}
if (n2g_idx) {
sprintf(n2g_idxF, "%s.n2g.idx.fa", refName);
fout.open(n2g_idxF);
for (int i = 1; i <= M; ++i)
fout<< '>'<< refs.getRef(i)->getName()<< endl<< n2g(refs.getRef(i)->getSeq())<< endl;
fout.close();
if (verbose) printf("%s is generated!\n", n2g_idxF);
}
}