averagehat
commented
6 years ago Copied from gitter:
The difference is more or less consistent with the output of bwa, which showed a similar discrepancy of about 460
it looks like c1 reported more reads mapped/unmapped than it had as actual input, which can happen due to chimeric alignment I guess the lesson here is that unmapped+mapped =/= total reads input, and if for some reason you want to know how many reads were input to bwa, you should look in the trimmed_reads directory http://seqanswers.com/forums/showthread.php?t=32708
Run setting: default configuration running 1 sample against 2 different references (total number of reads did not match). For troubleshooting purposes, ngs_mapper was run twice, 1 sample against 1 reference.
Since it is the same sample, the total number of reads from both results should be the same. The runs produced different total reads: 14999_c1 has 361350 reads 14999_c2 has 360988 reads Only difference should be the number of mapped and unmapped reads.
Attached below are the reference files, qualdepth.pdf files, and json files:
14999_references.zip 14999_c1_stats.zip 14999_c2_stats.zip