PCR duplicate removal for whole genome sequencing vs. whole exome sequencing

[cliff]

Hi

I did whole-genome sequencing and whole-exome sequencing on a whole-genome amplified (WGA’d) sample and got 2% of reads removed as duplicates by whole-genome sequencing but 80% of reads removed by whole-exome sequencing.

I then did whole-exome sequencing on an unamplified HapMap control and WGA’d HapMap control and got 25% of reads removed from the unamplified HapMap control and 50% removed from the WGA’d HapMap control.

I used Illumina standard PE101 whole-genome sequencing protocol for whole-genome sequencing and NimbleGen exome capture (version 2) for exome capture followed by Illumina sequencing.

Can anyone share some thoughts on the big difference between whole-genome sequencing and whole-exome sequencing of my WGA’d sample in terms of duplicate removal? All your comments will be greatly appreciated!

[lletourn]

We had the same issue at the beggining, hitting >60% dupes. We had to start with more dna and use a bit bigger fragment lengths to lower the values. We now typically get 20-30% dups.

Lets not forget that if you have 100x coverage with 100bp reads the chances of having 2 reads with the same 5' position and/or having 2 fragments having the same sequence is pretty high. So at high coverage many duplicates aren't duplicates.

How much coverage are you getting?