BLASR adds " RQ=" to read name when unmapped

[ryananeff]

From BamUtils, I am seeing many INVALID_QNAMEs for unmapped reads (example below):

Number of records read = 534176
Number of valid records = 508031

Error Counts:
INVALID_QNAME: 26145

TotalReads 534176.00
MappedReads 508031.00
PairedReads 0.00
ProperPair 0.00
DuplicateReads 0.00
QCFailureReads 0.00

MappingRate(%) 95.11
PairedReads(%) 0.00
ProperPair(%) 0.00
DupRate(%) 0.00
QCFailRate(%) 0.00

TotalBases 4711273300.00
BasesInMappedReads 4615088843.00
Returning: 7 (INVALID)

When I look at mapped reads vs. unmapped, I am seeing this difference:

Mapped:
m150224_063755_42156_c100788292550000001823173308251556_s1_p0/133171/0_211 10 007195F-001-01 ...

Unmapped:
m150225_013529_42156_c100779742550000001823166508251511_s1_p0/11992/37280_42077 RQ=0.787/0_4797 4 * ....

I think to fix this for the files I have right now is to convert back to SAM, strip out the RQ columns, and re-make the BAM

[mchaisso]

Are you using blasr from this repo or the pacbio one?

[ryananeff]

This repo. Was this fixed in master?

[mchaisso]

No, but blasr does not attach the RQ field to the read name and does not
have a problem from bax.h5 files. Are you aligning fasta files that have
this? In any case, I pushed a fix for this issue. Let me know if it works
for you.

-mark

On Tue, Jul 7, 2015 at 7:21 AM, Ryan Neff notifications@github.com wrote:

This repo. Was this fixed in master?

—
Reply to this email directly or view it on GitHub
#3 (comment).

[ryananeff]

This was from fasta files, hot h5. Anyways, I'll try out the fix with the next commit shows up and let you know how it goes!