void do_templates ()
{
int i;
GLOBAL = new_stream ();
tpls = (Template**) alloca (c_nsym * sizeof (Template*));
for (i = 0; i < c_nsym; i++)
tpls [i] = 0;
pass ();
for (i = 0; i < c_nsym; i++)
if (tpls [i]) {
free (tpls [i]->BODY);
free (tpls [i]);
}
free (CODE);
CODE = combine_output (GLOBAL);
}
////////////////////////////////////////////////////////////
// main
////////////////////////////////////////////////////////////
int main(int argc, char **argv) {
parse_command_line(argc, argv);
// prepare AT and GC counts
unsigned long long atgc[2] = {0};
// make trusted kmer data structure
bithash *trusted = new bithash(k);
// get good kmers from Hammer
if (hammerf != NULL) {
string hammerf_str(hammerf);
if (hammerf_str.substr(hammerf_str.size()-3) == ".gz") {
igzstream hammerf_in(hammerf);
trusted->hammer_file_load(hammerf_in, atgc);
} else {
ifstream hammerf_in(hammerf);
trusted->hammer_file_load(hammerf_in, atgc);
}
}
// get kmer counts
if(merf != NULL) {
string merf_str(merf);
if(ATcutf != NULL) {
if(merf_str.substr(merf_str.size()-3) == ".gz") {
igzstream mer_in(merf);
trusted->tab_file_load(mer_in, load_AT_cutoffs(), atgc);
} else {
ifstream mer_in(merf);
trusted->tab_file_load(mer_in, load_AT_cutoffs(), atgc);
}
} else {
if(merf_str.substr(merf_str.size()-3) == ".gz") {
igzstream mer_in(merf);
trusted->tab_file_load(mer_in, cutoff, atgc);
} else {
ifstream mer_in(merf);
trusted->tab_file_load(mer_in, cutoff, atgc);
}
}
// saved bithash
} else if(bithashf != NULL) {
if(strcmp(bithashf,"-") == 0) {
cerr << "Saved bithash cannot be piped in. Please specify file." << endl;
exit(EXIT_FAILURE);
} else
trusted->binary_file_input(bithashf, atgc);
}
cout << trusted->num_kmers() << " trusted kmers" << endl;
double prior_prob[4];
prior_prob[0] = (double)atgc[0] / (double)(atgc[0]+atgc[1]) / 2.0;
prior_prob[1] = .5 - prior_prob[0];
prior_prob[2] = prior_prob[1];
prior_prob[3] = prior_prob[0];
//cout << "AT: " << atgc[0] << " GC: " << atgc[1] << endl;
cout << "AT% = " << (2*prior_prob[0]) << endl;
// make list of files
vector<string> fastqfs;
vector<int> pairedend_codes;
parse_fastq(fastqfs, pairedend_codes);
// process each file
string fqf;
bool zip;
for(int f = 0; f < fastqfs.size(); f++) {
fqf = fastqfs[f];
cout << fqf << endl;
// unzip
if(fqf.substr(fqf.size()-3) == ".gz") {
zip = true;
unzip_fastq(fqf);
} else
zip = false;
// determine quality value scale
if(Read::quality_scale == -1)
guess_quality_scale(fqf);
// split file
vector<streampos> starts;
vector<unsigned long long> counts;
chunkify_fastq(fqf, starts, counts);
// learn nt->nt transitions
double ntnt_prob[Read::max_qual][4][4] = {0};
for(int q = 0; q < Read::max_qual; q++)
for(int i = 0; i < 4; i++)
for(int j = 0; j < 4; j++)
if(i != j)
ntnt_prob[q][i][j] = 1.0/3.0;
if(!TESTING)
learn_errors(fqf, trusted, starts, counts, ntnt_prob, prior_prob);
// correct
correct_reads(fqf, pairedend_codes[f], trusted, starts, counts, ntnt_prob, prior_prob);
// combine
if(pairedend_codes[f] == 0) {
combine_output(fqf, string("cor"), uncorrected_out);
}
// combine paired end
if(pairedend_codes[f] == 2) {
if(!zip) {
combine_output_paired(fastqfs[f-1], fqf, string("cor"), uncorrected_out);
} else {
combine_output_paired(fastqfs[f-1].substr(0,fastqfs[f-1].size()-3), fqf, string("cor"), uncorrected_out);
}
}
if(zip)
zip_fastq(fqf);
}
return 0;
}
