Convert between formats for names of indels

Hi,

I have a list of INDELS in their rsid format, and I am trying to convert this list from the rsids to a format like the one coming out of the imputation from MACH/Minimac, i.e. chr:pos:ALLELES.

I have tried using biomaRt to find the chr, positions, and alleles corresponding to the rsids like this:

library(biomaRt)
snpmart = useMart("snp", dataset="hsapiens_snp")
getBM(c("refsnp_id","allele","chr_name","chrom_start"), values="rs200623867",   filters="snp_filter", mart=snpmart)

But I get this for an INSERTION:

  refsnp_id   allele chr_name chrom_start
rs146107628      -/T       10   100002842

That I would like to convert to this format:

10:100002841:C_CT    I    R

While for a DELETION:

rs200623867    G/-       10   100003302

That I would like to convert to this:

10:100003301:AG_A    D    R

So it looks like I am missing the information about the other allele when using biomaRt.

Is there maybe a better approach to completing this convertion in R?

Thank you!

Simone

indels r biomart

• 3,445 views • link

updated 4.6 years ago by Ram

• written 11.9 years ago by lillo.sim •

1 answer

You could get 1 base upstream sequence with biomaRt

• 0 views • link

updated 4.6 years ago by Ram

• written 11.9 years ago by Emily

Thank you for your reply! But even if I go one base before I am still having trouble with finding the "other" allele using biomaRt, so for example how would I know to convert

rs146107628      -/T       10   100002842 --> 10:100002841:C_CT    I    R

If I don't know that the other allele is C, but I only get T from biomaRt?

• 1 views • link

updated 4.6 years ago by Ram

• written 11.9 years ago by lillo.sim •

Well, if you get one base upstream, that base would be C, so you would say -/T with one base upstream was C_CT.

• 0 views • link

written 11.9 years ago by Emily

Thanks Emily, this seems very clever. Do you know if there is a way using biomaRt to request the upstream position based on rsids in one go, or do you think I first need to find the chr:pos for each of the variants listed in the rsids object, and then in a second step find the alleles at their chr:pos-1?

I am just trying to understand the quickest way to convert these IDs inside an R script, maybe biomaRt isn't the best way to go if it requires many steps, but I don't know of another package that could help with this...

Thanks again for your advice!

• 0 views • link

written 11.9 years ago by lillo.sim •

In the variant mart, there's a section called Sequences where you can specify the upstream sequences plus the alleles.

• 0 views • link

written 11.9 years ago by Emily