Feature request: add query length to kmer_conservation output

[cschu]

Hi,

I'd be interested in having fulgor report the length of the query in kmer_conservation in order to facilitate assessment of how well the query matches. (E.g, 70 shared 31mers are a great match for 100bp, but less so for a 150bp or 250bp read.) My input are preprocessed WGS/RNAseq reads of varying lengths, so I cannot work with a uniform query length and parsing the input fastqs to count bases for all reads ist not really efficient, especially if fulgor already has that information.

I have a customised fulgor version that already does this, but I thought it could be of use for others and the code changes are trivial. The output for the non-matching reads is theoretically not necessary, but leaving it unchanged would screw with the output format.

Thanks,
Christian

diff --git a/tools/kmer_conservation.cpp b/tools/kmer_conservation.cpp
index 1851bdf..6eb561f 100644
--- a/tools/kmer_conservation.cpp
+++ b/tools/kmer_conservation.cpp
@@ -24,14 +24,14 @@ void kmer_conservation(FulgorIndex const& index,
             index.kmer_conservation(record.seq, kmer_conservation_info);
             buff_size += 1;
             if (!kmer_conservation_info.empty()) {
-                ss << record.name << '\t' << kmer_conservation_info.size();
+                ss << record.name << '\t' << record.seq.length() << '\t' << kmer_conservation_info.size();
                 for (auto kct : kmer_conservation_info) {
                     ss << "\t(" << kct.start_pos_in_query << ' ' << kct.num_kmers << ' '
                        << kct.color_set_id << ')';
                 }
                 ss << '\n';
             } else {
-                ss << record.name << "\t0\n";
+                ss << record.name << '\t' << record.seq.length() << "\t0\n";
             }
             kmer_conservation_info.clear();
             options.increment_processed_reads();

[jermp]

Hi @cschu and thank you for you interest in Fulgor!
Yes, you suggestion make sense.

I'm on vacation right now but will come back to this at the end of the month.

Best,
-Giulio