Hi I am not very clear about fields in BQSR table file(like errors,EmpiricalQuality...).I have tried finding answers,but the result is not very ideal.Could you give me some advice?
Have a look at this thread for an explanation of errors. The accuracy is simply the empirical quality score (what is calculated by the tool) - reported quality score. You may also find the BQSR presentations helpful. You can find them in the Presentations section.
Hi @Ruler. Please see Article#44 for a conceptual overview of BQSR. Namely, BQSR correlates basecall features with (i) read group sample (per-lane, per-sample), (ii) reported base quality score, (iii) position with the read (machine cycle) and (iv) sequence context (e.g. di and tri-nucleotide). It calculates the error empirically to then generate accurate base substitution, insertion and deletion quality scores. Unfortunately, beyond this conceptual overview, I haven't studied the reports that BQSR generates to know what each field represents. Are you running into errors?
Thanks for you answer.I have read the article before and so I simply know the conceptual overview of BQSR.But when I see the BQSR table and AnalyzeCovariates csv file,I always can't explain all fields,expecially Errors field in BQSR table and AnalyzeCovariates csv file and Accuracy field in AnalyzeCovariates csv file.
This is a picture of AnalyzeCovariates table:
And I am not running into errors,it's ok.
I've asked others on the team to followup, given I'm not familiar myself with these. They should get to your question soon.
OK @shlee.Thanks for your help and I am looking forward to the answer from others on your team.