BED file for WES data

[mybabyfox]

Dear ASCAT authors,

Thank you very much for contributing such a useful tool for the field.

I have a typical paired tumor-nomal WES dataset. I performed the standard ASCAT run with a provided BED for sequenced regions. And I noticed that the result is slightly different for a few regions if I remove the BED file from ascat.prepareHTS().

I know that it is recommended that a BED file is provided. Is that because ASCAT can avoid using the low-coverage off-target regions?

Thank you very much!

[tlesluyes]

Hi @mybabyfox,

Thank you for your feedback :)

That's correct: when running ascat.prepareHTS, one can specify a BED file that is used to focus on targeted regions (the BED_file argument), otherwise ASCAT will consider all SNPs covered by at least minCounts (in the germline). As you mentioned, we recommend providing a BED file so ASCAT can ignore off-target (= likely noisy) regions. Therefore, the slight difference between your two runs is likely due to noise in off-target regions and extra CNAs should be treated as artefacts.

One possible (minor) improvement would be to extend the BED regions by 500bp (both downstream and upstream) because the targeted experiment often sequences a bit more than expected and this could cover extra SNPs nearby (typically intronic SNPs next to exons covered by the sequencing design), slightly improving the resolution. This is not mentioned in the doc because one could take it for granted, although this may not necessarily be true for all sequencing experiments.

Cheers,

Tom.

[mybabyfox]

Hi Tom,

Your prompt response is really appreciated. And your answer makes great sense!

Best, Vida