SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
more precision about VarScan somatic Jane M Bioinformatics 35 09-14-2017 02:59 PM
VarScan output to VCF OlgaMikh Bioinformatics 4 06-07-2012 11:46 PM
varscan output jgSoton Bioinformatics 1 12-20-2011 08:39 AM
Varscan and mpileup aunderwo Bioinformatics 2 11-05-2011 03:43 AM
error from VarScan shuang Bioinformatics 1 10-13-2011 07:35 AM

Reply
 
Thread Tools
Old 02-12-2012, 10:03 PM   #1
alexbmp
Member
 
Location: Seoul, South Korea

Join Date: Oct 2011
Posts: 30
Default Questions about VarScan

Hi all. I'm a student working to identify tumor-specific mutations.

Recently I've came up to VarScan and SomaticSniper.

There are some confusions about using VarScan, and I see that the documentation is somewhat lacking in catching up its own versions, so I ask this question here !

I'm using VarScan.v2.2.8.jar somatic when doing my analyses.

1. When What is the "germline p-value" exactly?
1.1. I see that --somatic-p-value sets a p-value to statistically test (Fisher's exact test) whether "normal variant allele frequency of one locus" is the same as "tumor variant allele frequency of one locus".

( Many thanks about the discussion here (it really was a great help! ):
http://seqanswers.com/forums/showthr...hlight=varscan )

So for example, I thought such contingency table could be made (I drew it crudely because I could not put in spaces instead of hyphens and underscores; ugly, isn't it?):
------------Normal--Tumor
Reference___10______2
1 Variant_____2______8 --> Fisher's exact p-value calculated.

...where the counts are reads mapped to the locus.

1.2. However, how can a "germline p-value" be calculated? D. Koboldt (the developer) himself explained that this germline p-value is "the null hypothesis is that the sample is homozygous-reference; under this hypothesis, all reads should support the reference base", which I do not understand exactly. I ask for a comment about this matter.


2. Can VarScan emit output in VCF (variant call format)?
It is written that now VarScan supports output to be in VCF, as written on the main web page as "VarScan v2.2.8 released with new somatic calling features: Tumor-normal mpileup compatibility and VCF 4.1 output option." (http://varscan.sourceforge.net/index.html)
However, all I can find inside the documentation (or I'm doing something wrong) is about v2.2, not v2.2.8.


I really hope somebody could give me a warm hand.

Have a great day!!

Last edited by alexbmp; 02-12-2012 at 10:13 PM. Reason: To draw a table and add details
alexbmp is offline   Reply With Quote
Old 02-13-2012, 01:34 PM   #2
dkoboldt
Member
 
Location: St. Louis

Join Date: Mar 2009
Posts: 62
Default

Hello,

The germline P-value in VarScan output is computed when a site has been classified as Germline (the same in both normal and tumor). In this situation, the read counts from normal and tumor are combined into a total number for each allele (reference and variant). This is compared via Fisher's Exact Test to the null hypothesis, a wild-type position, at which all reads should reflect the reference allele.

The VCF 4.1 output option is available with VarScan 2.2.8 and above. However, it is currently only enabled for multi-sample calling functions mpileup2snp, mpileup2indel, and mpileup2cns. We are working on making VCF output an option for somatic mutation calling for the next release.
dkoboldt is offline   Reply With Quote
Old 08-02-2012, 06:54 AM   #3
blackgore
Member
 
Location: UK

Join Date: Sep 2009
Posts: 19
Default

Hi dkoboldt,
I've just started using VarScan (v2.2.11) and so far am very happy with what it can do. I am formatting my output from mpilup2snp and mpileup2indel as VCF, but, for me at least, the output is in version 4.0 (see header information below), not 4.1 as stated above. Is there another option distinct from setting "--output-vcf 1" to set it to version 4.1?

##fileformat=VCFv4.0
##source=VarScan2
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Depth">
##FILTER=<ID=str10,Description="Less than 10% or more than 90% of variant supporting reads on one strand">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
blackgore is offline   Reply With Quote
Old 08-16-2012, 10:18 AM   #4
dkoboldt
Member
 
Location: St. Louis

Join Date: Mar 2009
Posts: 62
Default

Hello,

We've just released VarScan v2.3.1 with better VCF compatibility; it should create VCF4.1 output. If it does not, please let me know!

Yours,

Dan Koboldt
dkoboldt is offline   Reply With Quote
Old 09-04-2012, 05:32 AM   #5
blackgore
Member
 
Location: UK

Join Date: Sep 2009
Posts: 19
Default

Hi Dan,
thanks very much for the announcement, and the new format seems to be working great!
Gordon
blackgore is offline   Reply With Quote
Old 02-20-2014, 07:38 AM   #6
FrankiB
Member
 
Location: Sherbrooke

Join Date: Dec 2013
Posts: 23
Default

Hi,

Is it possible to get vcf output using somatic mutation calling? If yes what should be the option in the command line?
FrankiB is offline   Reply With Quote
Old 02-20-2014, 07:58 AM   #7
Bukowski
Senior Member
 
Location: Aberdeen, Scotland

Join Date: Jan 2010
Posts: 347
Default

Quote:
Originally Posted by FrankiB View Post
Hi,

Is it possible to get vcf output using somatic mutation calling? If yes what should be the option in the command line?
Same as for germline:

--output-vcf 1
Bukowski is offline   Reply With Quote
Old 02-20-2014, 10:05 AM   #8
FrankiB
Member
 
Location: Sherbrooke

Join Date: Dec 2013
Posts: 23
Default

Yes I see. Thank you very much
FrankiB is offline   Reply With Quote
Reply

Tags
varscan

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 09:22 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO