How do I get the RQ= information in the header of the PB fastq ou fasta files

[cklopp]

Hi,

I've extracted fastq and fasta from bas.h5 files using bash5tools.py, but the output file did not contain the RQ= information in the header.

What is the tools I should use to collect this information?

Christophe

[rafalwoycicki]

SMRT Analysis

Hi,
You should use SMRT Analysis for this. bash5tools.py is not giving this info. Ask your sequence provider to extract subreads for you using SMRT Analysis if you do not have it.

Cheers,
Rafal

[cklopp]

Thank you for your reply. How can I do this using a command line?

Cheers,
Christophe

[GenoMax]

@rhall from PacBio participates in the forum and he should be able to tell if this can be done on the command line. @cklopp: You may want to send Dr. Hall a PM with a link for this thread.

[rhall]

You could estimate the RQ from the base quality values in the fastq file. It wont be exactly the same, but would be close enough for any use case I can think of. Alternatively you can use pbcore library to write a simple conversion tool yourself to include the information in the output header https://github.com/PacificBiosciences/pbcore.
In reality the RQ values are not well calibrated, and above ~0.75 - 0.8 do not correlate well with actual accuracy. I generally filter at 0.75 or 0.8 RQ using bash5tools.py then forget about the RQ values. For raw reads the base quality values in the fastq are also generally of little use, without going through some recalibration. Note the fastq values in ccs data are a lot more meaningful.

[cklopp]

Thank you for you reply. When I compare the alignment quality versus the quality values found in the fasta ou fastq file, they seem to correlate quiet nicely.