void Genotype::readBIMFile(std::string f)
{
std::ifstream file;
std::string line;
if(communicator->mpiRoot)
{
misc.message << "Reading SNPs data from file [ " << f << " ] ..." << std::endl;
misc.checkFileExists(f);
file.open(f.c_str());
this->SNPs.clear();
this->SNPIds.clear();
this->SNPIdsIdx.clear();
int idx = 0;
while(getline(file,line))
{
if(!file)
{
break;
}
std::istringstream sstemp(line); //->Comprova que te 6 elements.
std::string temp[6];
sstemp >> temp[0] >> temp[1] >> temp[2] >> temp[3] >> temp[4] >> temp[5];
SNP snp = SNP(temp[0], temp[1], temp[3], temp[4], temp[5]);
this->SNPs.push_back(snp);
this->SNPIds.push_back(snp.name);
if(this->SNPIdsIdx.count(snp.name) != 0)
{
misc.error("Error: The SNP with name: " + snp.name + " appears more than one time in the genotypes BIM file.", 0);
}
this->SNPIdsIdx[snp.name] = idx;
idx++;
}
file.close();
this->nSNPs = SNPs.size();
misc.message << this->nSNPs << " SNPs found." << std::endl;
}
// Addition on 29/12/13 (to be compatible with bowtie2):
// CIGAR_indels is field #5 in the sam format
// CIGAR_mismatches is the field that starts with MD:Z.
void ReadMapping::determine_snps(const DNAString& dna_str, const string& quality_str, string CIGAR_indels, string CIGAR_mismatches, size_t ref_pos)
{
if(CIGAR_mismatches.size() == 0)
return;
// This regexp was taken from the SAM manual
// /^MD:Z:([0-9]+)(([ACGTN]|\^[ACGTN]+)[0-9]+)*$/)
string::const_iterator p = CIGAR_mismatches.begin();
if((*p!='M') || (*(p+1)!='D') || (*(p+2)!=':') || (*(p+3)!='Z') || (*(p+4)!=':')) {
throw Illegal_mapping(CIGAR_mismatches, "Illegal SNPs description, expected MD:Z: at the beginning of CIGAR_mismatches format but found something different", __PRETTY_FUNCTION__);
}
p += 5;
// Read the offset to the first SNP
size_t read_pos = 0, ref_deleted = 0;
while((*p>='0') && (*p<='9')) {
read_pos *= 10;
read_pos += (*(p++)-'0');
}
// Set the offset to 0
read_pos--;
// Keep on reading the next snps until we reach the end
while(p != CIGAR_mismatches.end()) {
// First we expect the ref's char
if((*p!='^') && ((*p<'A') || (*p>'Z')))
throw Illegal_mapping(CIGAR_mismatches, string("Illegal SNPs description (2), expected CIGAR_mismatches format but found something different"), __PRETTY_FUNCTION__);
if(*p=='^') {
p++;
while(((*p<'0') || (*p>'9')) && (p != CIGAR_mismatches.end())) {
ref_deleted++;
m_snps.push_back(SNP(SNP::DELETION, ref_pos+read_pos+ref_deleted, read_pos, *p, '-', 40));
p++;
}
}
else {
read_pos++;
m_snps.push_back(SNP(SNP::MISMATCH, ref_pos+read_pos+ref_deleted, read_pos, *p, dna_str.get(read_pos), quality_str.at(read_pos)));
p++;
}
// There must be at least one more number
if((*p < '0') || (*p > '9'))
throw Illegal_mapping(CIGAR_mismatches, "Illegal SNPs description, expeected CIGAR_mismatches format but found something different", __PRETTY_FUNCTION__);
unsigned int o = 0;
while((*p>='0') && (*p<='9')) {
o *= 10;
o += (*(p++)-'0');
}
read_pos += o;
}
//cerr << endl << "--------------------------------------------------------------" << endl;
//for(vector<SNP>::iterator itt=m_snps.begin(); itt!=m_snps.end(); itt++) {
// cerr << (itt-m_snps.begin()) << '\t' << itt->ref_pos << '\t' << itt->read_pos << '\t' << (itt->snp_type==SNP::MISMATCH) << (itt->snp_type==SNP::INSERTION) << (itt->snp_type==SNP::DELETION) << endl;
//}
//cerr << endl << "+++" << endl << endl;
// Now we have to go through the insertion/deletion CIGAR string and correct mismatches accordingly
// Examples: 6M1D9M1I132M, 31M1I106M
p = CIGAR_indels.begin();
// Make sure that the end of the string is legal (so we don't have to check it all the time in the loop below)
char ch = *(CIGAR_indels.rbegin());
if((ch!='I') && (ch!='D') && (ch!='M')) {
throw Illegal_mapping(CIGAR_indels, string("Illegal CIGAR-indels description (3)"), __PRETTY_FUNCTION__);
}
int read_pos2 = 0;
size_t ref_pos2 = ref_pos;
while(p!=CIGAR_indels.end()) {
size_t o = 0;
while((*p >= '0') && (*p <= '9')) {
o *= 10;
o += (*(p++)-'0');
}
if(o == 0) {
throw Illegal_mapping(CIGAR_indels, string("Illegal CIGAR-indels description (4)"), __PRETTY_FUNCTION__);
}
vector<SNP>::iterator ps;
size_t pos;
switch (*(p++)) {
case 'M':
read_pos2 += o;
ref_pos2 += o;
break;
case 'D':
// Nothig to do here - this case was already taken care of in CIGAR_mismatches
ref_pos2 += o;
break;
case 'I':
for(ps=m_snps.begin(); ps!=m_snps.end(); ps++) {
if(ps->read_pos >= read_pos2) {
ps->read_pos += o;
}
}
read_pos += o;
for(ps=m_snps.begin(); (ps!=m_snps.end()) && (ps->read_pos < read_pos2); ps++)
;
// At this point we don't provide any information about the insetion except that it exists
m_snps.insert(ps, SNP(SNP::INSERTION, ref_pos2-1, read_pos2-1, '-', '-', '-'));
read_pos2 += o;
// ref_pos2 += o;
// pos = ps-m_snps.begin();
// while(o > 0) {
// ps = m_snps.begin()+pos;
// pos++;
// m_snps.insert(ps, SNP(SNP::INSERTION, ref_pos2++, read_pos2-1, '-', dna_str.get(read_pos2), quality_str.at(read_pos2)));
// o--;
// read_pos2++;
// }
break;
default:
throw Illegal_mapping(CIGAR_indels, string("Unexpected CIGAR-indels operation found"), __PRETTY_FUNCTION__);
}
}
// Sanity check
if(((read_pos+1) != dna_str.size()) || (read_pos2 != dna_str.size()))
throw Illegal_mapping(CIGAR_indels+string(", ")+CIGAR_mismatches, "snps and offsets do not sum to expected read length", __PRETTY_FUNCTION__);
//for(vector<SNP>::iterator itt=m_snps.begin(); itt!=m_snps.end(); itt++) {
// cerr << (itt-m_snps.begin()) << '\t' << itt->ref_pos << '\t' << itt->read_pos << '\t' << (itt->snp_type==SNP::MISMATCH) << (itt->snp_type==SNP::INSERTION) << (itt->snp_type==SNP::DELETION) << endl;
//}
}
