Clarification of the TAR and TIR FORMAT fields in Indel VCF file

[Fong Chun Chan]

Hi,

I noticed that in the indel VCF files there are two format fields:

##FORMAT=<ID=TAR,Number=2,Type=Integer,Description="Reads strongly supporting alternate allele for tiers 1,2">

##FORMAT=<ID=TIR,Number=2,Type=Integer,Description="Reads strongly supporting indel allele for tiers 1,2">

What exactly is the difference between the "alternative allele" and "indel allele" here? Is this potentially a typo and TAR should actually be supporting the "reference allele"? 

Thanks,

[Saunders, Chris]

Hi Fong,

 

Alternative allele means the reference allele in addition to any other conflicting/overlapping candidate indel alleles.

 

-Chris

--
You received this message because you are subscribed to the Google Groups "strelka-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to strelka-discu...@googlegroups.com.
To post to this group, send email to strelka...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/strelka-discuss/197d2b68-b7ce-4b60-b3c8-caa073457213%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Andrew Skol]

Hi Chris,
 I had the same question as Fong, and unfortunately I'm having difficulty understanding your response. Let me show you why I'm having problems:

Let's say we have the following line from the output:
#CHROM  POS     ID      REF     ALT     QUAL    FILTER  INFO    FORMAT  NORMAL  TUMOR
2       206550942       .       CCCT    C       .       QSI_ref IC=6;IHP=5;NT=ref;QSI=17;QSI_NT=17;RC=7;RU=CCT;SGT=ref->hom;SOMATIC;TQSI=1;TQSI_NT=1  DP:DP2:TAR:TIR:TOR:DP50:FDP50:SUBDP50   11:11:11,11:0,0:0,0:10.74:0.54:0.00   4:4:0,0:3,4:1,1:3.27:0.00:0.00

Here the reference allele is listed as "CCCT" and the alternative is "C".

If we restrict attention to tier 1 calls, then in the Normal sample,
1. Read depth is 11
2. Reads supporting the "alternative" allele is 11.
3. Reads supporting the "indel" allele is 0.
4. Other reads is also 0.

Thus since you say that "alternative" means reference and any other conflicting/overlapping candidate indel allele, the normal tissue has 11 instances in which "CCCT" was observed, and 0 in which "C" was observered. So this consistent with your definition above.

However. . .
In the Tumor sample tier 1 calls.
1. Depth is 4.
2. Reads supporting the "alternative" allele is 0.
3. Reads supporting the "indel" allele is 3.
4. Other reads is also 1.

Thus in this the tumor we see no instances of the reference allele "CCCT" since reads supporting the "alternative" allele is 0, and we see 3 copies of the alternative allele "C" and 1 instance of another read.

But if what you said above is correct: "Alternative allele means the reference allele in addition to any other conflicting/overlapping candidate indel alleles", then why is TAR not 1, since there was 1 "other" read?

Thanks for your help with this.

Andrew

[Saunders, Chris]

Hi Andrew,

 

The “other” category means that among the set of indel alleles considered, the posterior probability of any single indel allele conditioned on the read observation is not high (in this case meaning < Q30). The principal is that the read has to uniquely support one allele, so a read which for instance just crosses into a long span of breakend homology is not going to meet this criteria. In this case the “set of indel alleles considered” includes the indel allele, the reference allele, and zero to several overlapping alternate indel alleles (the great majority of the time this will be zero). Hope that helps.

 

-Chris

To view this discussion on the web visit https://groups.google.com/d/msgid/strelka-discuss/53db227f-00e9-4eef-88c5-0c57f2fee11d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.