unsigned long int fq_stream_trimmer(UT_string *fq_fn, int pipe_fd, UT_string *out_prefix, int no_pre, int len_pre, unsigned long int *comp_cnt, unsigned long int *org, char split, int fmt_fasta) {
UT_string *new_head_data;
utstring_new(new_head_data);
UT_string *head_data;
utstring_new(head_data);
UT_string *seq_data;
utstring_new(seq_data);
UT_string *extra_data;
utstring_new(extra_data);
UT_string *qual_data;
utstring_new(qual_data);
unsigned long int cnt = 0;
char *start = NULL;
char *end = NULL;
char *suffix = NULL;
FILE *fq_file = NULL;
FILE *pipe_in = fdopen(pipe_fd, "w");
if (!(utstring_len(fq_fn))) {
fclose(pipe_in);
return(0);
}
// Try to open the fastq file
if (!(fq_file = gzopen(utstring_body(fq_fn), "r"))) {
utstring_printf(fq_fn, ".gz");
if (!(fq_file = gzopen(utstring_body(fq_fn), "r"))) {
fclose(pipe_in);
return(0);
}
}
int x = 0;
char head_char = '@';
if (fmt_fasta) {
head_char = '>';
}
while (ss_gzget_utstring(fq_file, head_data)) {
ss_gzget_utstring(fq_file, seq_data);
ss_gzget_utstring(fq_file, extra_data);
ss_gzget_utstring(fq_file, qual_data);
if (!split || ((suffix = strchr(utstring_body(head_data), '/')) && (suffix[1] == split))) {
(*org)++;
if ((x = trimmer(utstring_body(qual_data))) >= min_len) { // Keep at least some of read
// Reject read if complexity is too low
if ((entropy_cutoff < 0.0) || (entropy_calc(utstring_body(seq_data), x) >= entropy_cutoff)) {
// Truncate sequence
ss_trim_utstring(seq_data, x);
ss_strcat_utstring(seq_data, "\n");
if (!fmt_fasta) {
ss_trim_utstring(qual_data, x);
ss_strcat_utstring(qual_data, "\n");
}
// Fixup the read name
utstring_clear(new_head_data);
end = strchr(utstring_body(head_data), ':');
if (no_pre) {
if ((start = strchr(utstring_body(head_data), '|'))) {
start++;
} else {
if (colorspace_flag) {
start = utstring_body(head_data) + 4;
} else {
start = utstring_body(head_data) + 1;
}
}
*end = '\0';
} else {
start = utstring_body(out_prefix);
}
end++;
if (colorspace_flag) {
if (len_pre) {
utstring_printf(new_head_data, "%c%.2s+%u|%s:%s",head_char,utstring_body(head_data)+1,x,start,end);
} else {
utstring_printf(new_head_data, "%c%.2s+%s:%s",head_char,utstring_body(head_data)+1,start,end);
}
} else {
if (len_pre) {
utstring_printf(new_head_data, "%c%u|%s:%s",head_char,x,start,end);
} else {
utstring_printf(new_head_data, "%c%s:%s",head_char,start,end);
}
}
fputs(utstring_body(new_head_data), pipe_in);
fputs(utstring_body(seq_data), pipe_in);
if (!fmt_fasta) {
fputs(utstring_body(extra_data), pipe_in);
fputs(utstring_body(qual_data), pipe_in);
}
cnt++;
} else {
// rejected by entropy filter
// Send along placeholder read to be discarded, keeping read1 and read2 in sync
// Empty fastq header is a read to be rejected by consumer threads
(*comp_cnt)++;
fputs("\n", pipe_in);
}
} else {
// rejected by minimum length cutoff
// Send along placeholder read to be discarded, keeping read1 and read2 in sync
// Empty fastq header is a read to be rejected by consumer threads
fputs("\n", pipe_in);
}
}
}
fclose(pipe_in);
gzclose(fq_file);
utstring_free(new_head_data);
utstring_free(head_data);
utstring_free(seq_data);
utstring_free(extra_data);
utstring_free(qual_data);
return(cnt);
}
int main(int argc, char* argv[])
{
Getopt getopt;
getopt.addToHelp(" <input file>");
if (getopt.processOpts(argc, argv)) {
getopt.showHelp(std::cout);
exit(1);
}
if(argc - getopt.first_non_opt() != 1) {
getopt.showHelp(std::cout);
exit(1);
}
const char* filename = argv[getopt.first_non_opt()];
// read data, create System pointer
printf("reading input file %s\n", filename);
LinAlg::System::ptr genes;
try {
Cq::CqFile cqfile(filename);
genes = Cq::read_data(cqfile);
}
catch(std::exception& e) {
printf("%s\n", e.what());
return 1;
}
Stat::PTree tree;
printf("adding vectors to tree\n");
tree.add_vectors(*genes);
printf("calculating probabilities\n");
tree.calc_probs();
// printf("running show_probs\n");
// tree.show_probs(stdout);
double total_prob = 0.0;
double entropy = 0.0;
for(size_t i = 0; i < genes->n_vectors(); ++i) {
const LinAlg::ConstVector cv(i, *genes);
const double prob = tree.vector_prob(cv);
total_prob += prob;
entropy += prob * log(prob) * -1.0;
}
printf("total prob=%f\n", total_prob);
printf("exp(entropy)=%f\n", exp(entropy));
printf("quitting\n");
return 0;
#if 0
const size_t nSymbols = LinAlg::count_letters(*genes);
Stat::EntropyCalculator entropy_calc(nSymbols, genes);
const double entropy = entropy_calc.entropy();
printf("entropy for %d genes: %f\n", (int)genes->n_vectors(), entropy);
const double should_be_one = entropy_calc.prob_check();
printf("should be 1 => %f\n", should_be_one);
return 0;
#endif
}
