WES - how deep do you sequence your germline control

onconaut · 02-12-2015, 09:28 AM

Hello Seqers,

we currently do WES with a mean coverage of ~120x for tumor as well as the germline samples. Since we only use the germline to help us with the somatic alteration calls I was thinking about cutting the coverage of the germline samples in half to free up some real estate on the flowcell. Does any one have experience with this?

Thanks, O.

Bukowski · 02-12-2015, 12:45 PM

We'd normally do 50x on the normal and 100x+ on the tumour - I think that's quite a common strategy.

bilyl · 02-12-2015, 01:55 PM

It depends on your downstream processes. If you are simply calling SNVs then it doesn't really matter. But for other applications (eg. CNVs), many programs prefer to have similar coverage between your tumor and normal. Even if they don't require it, you are getting rid of other sources of variation (eg. by capturing the same amount of DNA in the same exome hyb reaction).

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Finding CNVs in tumors, no control, WES - Which programs should I use?	Miau	Bioinformatics	0	08-01-2014 03:06 AM
NGS whole genome sequence versus sequence capture for quality control	houkto	Bioinformatics	0	02-02-2012 05:16 AM
PubMed: Control-free calling of copy number alterations in deep-sequencing data using	Newsbot!	Literature Watch	0	04-08-2011 02:10 AM
Sequence Alignment Quality Control	nilshomer	Bioinformatics	4	02-06-2010 08:24 AM

02-12-2015, 09:28 AM	#1
onconaut Junior Member Location: South Central US Join Date: Aug 2012 Posts: 7	WES - how deep do you sequence your germline control Hello Seqers, we currently do WES with a mean coverage of ~120x for tumor as well as the germline samples. Since we only use the germline to help us with the somatic alteration calls I was thinking about cutting the coverage of the germline samples in half to free up some real estate on the flowcell. Does any one have experience with this? Thanks, O.

02-12-2015, 12:45 PM	#2
Bukowski Senior Member Location: UK Join Date: Jan 2010 Posts: 390	We'd normally do 50x on the normal and 100x+ on the tumour - I think that's quite a common strategy.

02-12-2015, 01:55 PM	#3
bilyl Member Location: USA Join Date: Aug 2013 Posts: 52	It depends on your downstream processes. If you are simply calling SNVs then it doesn't really matter. But for other applications (eg. CNVs), many programs prefer to have similar coverage between your tumor and normal. Even if they don't require it, you are getting rid of other sources of variation (eg. by capturing the same amount of DNA in the same exome hyb reaction).