Which variant is correct?

Question

[michael]

As I mentioned in my first post yesterday I'm new at all this. (My interest is the Y and genetic genealogy.) Please be patient. :)

I'm looking at position 2880955. The first record below yields an A, the second a T. The ancestral value is C. Both MAPQ and QUAL are high scores. From what I understand, these should be sound reads. Am I misinterpreting something or is this to be expected? Does it boil down to consensus?

SN7001368:430:H2V2JADXX:2:2203:10989:33132 2:N:0:TAAGGCGA+AAGGAGTA      163
chrY    2880949 37      6=1X158=        =       2881012 228
CTGTCAAACAGCCCTAAGAGGTTCTGAGAACATGTGCCCAAGGTGAGCAGGGTACAGTCTGGTTT
BBBFFFFFFFFFIIIIIIIIIFIIIIIIIIIIIIIIIIIIIIIFFIIIIIIIBFIIIFIIIIFFI

SN7001368:430:H2V2JADXX:1:1215:16879:4926 2:N:0:TAAGGCGA+AAGGAGTA       163
chrY    2880951 38      4=1X78= =       2880951 83
  GTCATACAGCCCTCAAGAGGTTCGAGAACATGTGCCCAAGGTGAGCAGGGTACAGTCTGGTTT
  BBBFFFFFFFFFFIIIIIIFFIIIIIIIIIIIIIIIIIIIFBFFIIIIIIFFIIIIIIIIFFI

Thanks, Michael

Answer 1

[Persistent LABS]

Hi

First of all You need to check how many reads align at 2880955. Is it only 2 reads, as you have shown? If only 2, then its a low coverage position (Read depth=2). So simply based on read count information (50-50 in your case), it is very difficult to say whether its C -> A or C->T variant. For such a low coverage region, using prior information would better predict it as either variant or not. Use variant caller tools such as bcftools or GATK, which use probabilistic method (Bayesian) to infer variant rather.

Thanks

Priyabrata

Persistent LABS

[michael]

Thanks so much, Priyabrata. There are four reads, the last column being the start position of the strand, each appearing twice:

2880955 A 163 35 F 2880949 2880955 A 163 37 F 2880949 2880955 T 163 38 F 2880951 2880955 T 83 38 F 2880951

So, it seems that I'm reading it right but should consider the results unreliable, as opposed to those cases where I have a read depth of, say 20, with 19 Ts and 1 A.

Thanks! Michael