Upstream pseudogene causing MAPQ 0 and exclusion during variant calling

Question

[Joel Wallenius]

Hello,

just upstream of GBA1 is a pseudogene that is quite similar to GBA1, this makes most of my mapped reads have MAPQ 0 across this region (I think? I use bwa-mem2 with default settings), and the variants are not listed in the corresponding VCF file.

Can I remedy this somehow? Mask the pseudogene and map again?

Sincerely, Joel

[ATpoint]

Technically you can mask it, but what if the variant comes from the pseudogene and not the gene you "want"?

[Joel Wallenius]

Yes, well... surely it's better with some FPs and TPs than nothing at all, eh. Is there another way to do this? Hmm...

Answer 1

[Joel Wallenius]

Hey, I found the solution: Use a hg38 reference that does not include alt contigs, e.g.: https://ftp.ncbi.nlm.nih.gov/genomes/all/GCA/000/001/405/GCA_000001405.27_GRCh38.p12/GRCh38_major_release_seqs_for_alignment_pipelines/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna.gz

Don't ask me about the 000/001/405 parent folders. I have no idea what they mean. See here for more(?) info: https://lh3.github.io/2017/11/13/which-human-reference-genome-to-use

[GenoMax]

Don't ask me about the 000/001/405 parent folders. I have no idea what they mean.

They are simply parts of the accession number GCA_000 / 001 / 405 used to create a file system hierarchy.