search

epi2me-labs / wf-human-variation

Other
86 stars 41 forks source link

What's new in sniffles-2.0.7-epime? #170

Closed ymcki closed 2 months ago

ymcki commented 2 months ago

I find that the sniffles in the docker image ontresearch/wf-human-variation-sv:sha69e7a43c6f21c0ea2083e9719257fb44bc35c2ba referenced by 2.1.0 has a version of 2.0.7-epime

In the previous image referenced by earlier pipeline ontresearch/wf-human-variation-sv:shabc3ac908a14705f248cdf49f218956ec33e93ef9 The version is 2.0.7

I presume 2.0.7-epime is a customized version. So how does it differ from the original 2.0.7???

SamStudio8 commented 2 months ago

Hi @ymcki - the change integrates the filterBam step into Sniffles2 itself, removing the need to generate a filtered BAM on disk before calling Sniffles. We will be making the suggestion to integrate the change upstream. We'll aim to make such changes more transparent in future.

ymcki commented 2 months ago

Thanks for your quick reply.

My understanding of the filterBam step is to get rid of both supplementary alignments (2048) and secondary alignments (256).

Theoretically speaking, removing the former can lead to loss of information as there are quite some number of reads that are concatenation of multiple true reads probably due to failure of read splitting algorithm. https://github.com/nanoporetech/dorado/issues/702

Or is it because sniffles-2.0.7 cannot handle supplementary alignments properly??