Test translation of specific genotype cases for 23andMe converter

[aneilbaboo]

Create one or more input 23andMe file which encodes several specific situations.

Test corresponding rows of the imputed VCF:

Expected cases:

• homozygous mutation
• heterozygous mutation
• Y chromosome mutation
• validate that the reference base is correct for each mutation
  • "Reference base of NC00001.1:g.123123 should be A"
  • "Reference base of NC00003.1:g.55555 should be T"
  • etc.

Unexpected cases:

• invalid rsId
• invalid chromosome number
• invalid chromosome location (e.g., negative number, too large number, string)
• missing genotype entry

[taltman]

Hi @aneilbaboo

I grabbed a pair of input (23&Me) and output (VCF) from the S3 buckets that my test make targets populate. I randomly spot-checked a number of the categories above, comparing not just the file entries against one another looking for conversion errors, but also checked against dbSNP to make sure that chromosome, position, rsID, gene annotation, and reference base information looks correct.

I didn't find any errors. The only oddity that I found is that Beagle seems to output an extra line for X chromosome variants when the subject is male. I'll debug that issue this week, along with adding some high-throughput sanity checking of the entire input & output files.

[aneilbaboo]

@taltman thanks for the update. So it's duplicating male X-linked variants?