nygenome/lancet

Input bam

Closed this issue · 5 comments

Hi,
When I use Lancet, I'm not sure about the bam. Could Lancet realign bam itself ? or I need realign by GATK ?
Best wishes

No need to realign the BAM. Lancet, like many modern variant callers, uses local-assembly to deal with soft-clipped reads and other mis-alignments.

I have this question too, because I saw that GATK software tools (v.2.7-4) were used for improving
alignments around indels (GATK IndelRealigner) and base quality recalibration
(GATK base quality recalibration tool) using recommended parameters in the first Lancet paper (Genome-wide somatic variant calling using localized colored de Bruijn graphs)

We did that to fairly compare results across multiple tools using the same BAM as input. Indel realignment is not a requirement for Lancet.

Can lancet output bam file containing all reads which were repositioned or realigned, or some evidences that support the SNV in a visualized way ? So I can show the SNV sites in IGV or other tools.

It was said that "Colored de Bruijn graphs can be easily exported and rendered to visualize variants of interest in graph space (Fig. 1), which can help in confirming a variant. This feature complements read alignment visualization tools such as the integrative genomics viewer (IGV)10 and provides another useful view into the data that support variant calling." in the first lancet paper. But, It's hard to understand the Colored de Bruijn graphs for non-expert bioinformaticians. If only some detailed explaination were added !