#16 2 high quality snps called with low posterior propbablity and lost

2.0
closed
nobody
None
2013-05-13
2013-04-22
lichun
No

The following two snps are called with low posterior propbablity (0.31 and 0.001) with atlas-snp2.

Pileup for these two snps are:

chrX 1790568 G 31 AAA..ca.,a,A.,,aaA,AA,,,.a.aaaa !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
chrX 834396 A 24 ,tTt.t,TT,,.t.,,T,..,c., !!!!!!!!!!!!!!!!!!!!!!!!

1 Attachments

Discussion

  • Danny Challis

    Danny Challis - 2013-04-24
    • status: open --> accepted
     
  • Danny Challis

    Danny Challis - 2013-04-25

    I have looks at these two sites in depth using the files you provided and found that while the read depth, base-quality and distance to the 3' end look fairly good, about half of the reads do not pass the neighboring quality standard (NQS: SNP base has QUAL>=20 and SNPs within 5bp have QUAL>=15). These reads failing the NQS indicates the SNPs may in fact be sequencing errors. The model appears to be behaving as designed.

    One of the advantages of Atlas2 is the ability to tune the sensitivity and specificity. If you are aiming to increase the call set's sensitivity you may lower the posterior probability cutoff (default 0.95) and/or the prior probability of error (default 0.1). However, such adjustments will increase the likelihood of calling more false positive SNPs.

    Please let us know if you have any more questions.

     
  • Danny Challis

    Danny Challis - 2013-05-13
    • status: accepted --> closed
     


Anonymous

Cancel  Add attachments





Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks