Alignment: how to deal with read mapped on decoy

Question

[quentin54520]

Hello all,

I made my alignments with a human reference containing decoys. I know how to remove reads that aligned up with decoys but what I don't know is: should I do it?

For the header it's the same thing, do I have to remove the @SQ lines corresponding to the decoys ?

I have some reads with the R1 on "normal" chromosome and R2 on decoy, it's usual to have this type of alignment?

Thanks a lot in advance 😀

[rpolicastro]

What type of sequencing data is this, what is the goal of your analysis, and what software are you currently using?

[quentin54520]

It's whole genome sequencing from a cohort of patients. The goal is to identify the genetic causes of their rare disease ( SNV, short indel or SV). for the software: i used Bwa mem with the reference hg38 no_alt_plus_hs38d1_analysis_set.

[berick]

trash them. you won't be able to annotate or to make any sense out of them till a new genome build which integrates them comes.