Filtering sequencing errors

[msl1y11]

Hi All,

I am looking for a tool that filters reads from SAM/BAM files where the overlapping sequence of a paired read is discordant i.e. one read contains a sequencing error. I have looked around and started to write something in Perl but I am not confident it works.

Many thanks,
Matt

[kmcarr]

Hi All,

I am looking for a tool that filters reads from SAM/BAM files where the overlapping sequence of a paired read is discordant i.e. one read contains a sequencing error. I have looked around and started to write something in Perl but I am not confident it works.

Many thanks,
Matt

How do you know which is the error and which is the truth?

There is a large number of existing tools to merge overlapping PE reads and each will have it's own method of dealing with mismatching bases which may give you some guidance. See this blog post for a list of some of them.

[msl1y11]

Thanks your reply kmcarr. I was planning to discard both reads of the pair, assuming the probability of a sequencing error arising is random this should not skew allele representation.

Cheers
Matt

[SNPsaurus]

I use SeqPrep for this purpose. Discordant alleles will give a much lower quality score in the overlapped sequence and can be filtered out. Not exactly what you are asking for, but it might help.

[gaihua559]

May I ask a question that it is necessary to filter the reads with low quality score before aligned to the reference by bowtie or BWA? Thanks a lot!