understanding ariant allele frequency for paired ended reads

Question

[robinycfang]

I am trying to understand how VAF is calculated under the senario of paired ended reads and overlapping regions of paired reads covering the SNV.

1. Are the sequences in read1 and read2 reverse complementary? Are both of the sequenced presented in the same orientation as the reference genome in BAM (case 2 below) or they just stay as whatever original orientation they have (case 1 below)? I am trying to find the mutant alleles in BAM, so wondering should I check both orientations.

A to G SNV:
is it
---------------A-------------
---------------G-->
       <-------C---------- 
or?
---------------A-------------
---------------G-->
       <-------G----------

1. When we compute total depth for a locus, if a pair of reads have overlaps that cover the locus, should we treat it as one count or two for the total depth. What if the mutant allele are present in the pair, for mutatnt allele depth, should we count 1 or 2?

Thanks!

[LauferVA]

try biostars handbook

Answer 1

[Jeremy]

Paired ends do not overlap. There is a known distance separating them. And yes, they run in opposite directions. See the following website from Illumina, especially the figure:

Paired-end Sequencing

[robinycfang]

I believe some of the paired ended reads overlap due to the short DNA fragments they are sequencing...