trio pipeline

Answers

• Sheila Broad InstituteMember, Broadie ✭✭✭✭✭

  June 2017

  @vittoria
  Hi Vittoria,

  You may find the Genotype Refinement Workflow useful. You can find more information in the presentations section and Methods and Algorithms section.

  -Sheila

  0 · Share on Twitter Share on Google+
• vittoria Member

  July 2017

  thank you sheila

  I understand that the first step is running

  java -jar GenomeAnalysisToolkit.jar -R human_g1k_v37_decoy.fasta -T CalculateGenotypePosteriors --supporting 1000G_phase3_v4_20130502.sites.vcf -ped trio.ped -V recalibratedVariants.vcf -o recalibratedVariants.postCGP.vcf

  I understand that recalibratedVariants.vcf combines the variants of the father, the mother and the child
  i.e. triorecalibratedVariants.vcf
  but
  I have produced the recalibratedVariants.vcf for father, mother and child separately

  how do I put them together
  fatherrecalibratedVariants.vcf, motherrecalibratedVariants.vcf childrecalibratedVariants.vcf -->
  triorecalibratedVariants.vcf

  moreover can I find .ped file I can modify with my data?

  thank you

  vittoria

  0 · Share on Twitter Share on Google+
• vittoria Member

  July 2017

  Dear Sheila, thank you

  I understand that I have to run something like

  java -jar GenomeAnalysisToolkit.jar -R human_g1k_v37_decoy.fasta -T CalculateGenotypePosteriors --supporting 1000G_phase3_v4_20130502.sites.vcf -ped trio.ped -V recalibratedVariants.vcf -o recalibratedVariants.postCGP.vcf

  I assume that recalibratedVariants.vcf contains variants of the father, of the mother and of the kid
  I have 3 recalibrated vcf files one for the father, one for the mother and one for the kid
  how do I put together the files?
  second:
  I do I get an example of trio.ped? I would like to modify one

  thank you very much

  vittoria

  0 · Share on Twitter Share on Google+

  Issue · Github

  July 2017 by shlee

  Issue Number

  2305

  State

  closed

  Last Updated

  August 2017

  Assignee

  Array

  Closed By

  chandrans

• shlee CambridgeMember, Broadie ✭✭✭✭✭

  July 2017 edited July 2017

  Hi @vittoria,

  Sheila is away at a workshop and can followup on your questions when she is back in Boston next week.

  In the meanwhile, this thread and this thread may be of interest.

  0 · Share on Twitter Share on Google+
• Sheila Broad InstituteMember, Broadie ✭✭✭✭✭

  July 2017

  @vittoria
  Hi Vittoria,

  For the trio VCF, we recommend running the GVCF workflow. However, if you must, you can simply use CombineVariants to combine the three VCFs into one VCF.

  For creating the ped file, have a look here.

  -Sheila

  0 · Share on Twitter Share on Google+
• vittoria Member

  September 2017

  Dear friends, I generated a joint vcf file for a trio but I have problems visualizing it with igv. where can I find a sample file to see how a correct file should appear in igv? where is the information in each variant concerning the identity (son, father or mather)? I do not understand how I see whether a mutation is present in the sono but not in the parents and so on
  thank you maria vittoria

  0 · Share on Twitter Share on Google+
• Sheila Broad InstituteMember, Broadie ✭✭✭✭✭

  September 2017

  @vittoria
  Hi Maria,

  I think you will find the Variant Discovery tutorial here helpful.

  -Sheila

  0 · Share on Twitter Share on Google+
• artitandon Member ✭✭

  February 2018

  I have a trio whole exome data set, based on the VQSR documentation it seems that 3 samples is too few to run with VQSR. Is that correct? What is the best option in this?

  0 · Share on Twitter Share on Google+
• Sheila Broad InstituteMember, Broadie ✭✭✭✭✭

  February 2018

  @artitandon
  Hi,

  Yes, 3 exome samples are not enough for VQSR. We recommend using at least 30 exome samples. You can try hard filtering.

  -Sheila

  0 · Share on Twitter Share on Google+
• MasMarius Member

  June 2018

  Hi,

  I am struggling to understand the best practice to analyze a trio WES as I want to do it with the new GATK4 tools. May you correct my method and suggest best options.. let s see;

  First, data pre-processing starting from fastq: bwa --> samtools sort --> picard duplicates and fixed mates. Once completed these steps I wanted to Realign and Recalibrate respectively using RealignTargetCreator and IndelRealigner with GATK with dbsnp and indel GOLD's and the last BaseRecalibrator for dbsnps, but i found

  https://gatkforums.broadinstitute.org/gatk/discussion/11455/realignertargetcreator-and-indelrealigner?

  where is said to skip those and go straight to Printreads with new tool ApplyBSQR for BQSR and call HaplotypeCaller....

  Here I am getting confused, should i go straight after picard step to HaplotypeCaller in ERC GVCF mode and then use GenomicDBImport to pass the GenotypeGVCFs for joint calling? Otherwise what would be the best practice.

  As i also read https://gatkforums.broadinstitute.org/gatk/discussion/3893/calling-variants-on-cohorts-of-samples-using-the-haplotypecaller-in-gvcf-mode but cant get to merge in my head the best way.
  thank you in advance!!

  0 · Share on Twitter Share on Google+
• Sheila Broad InstituteMember, Broadie ✭✭✭✭✭

  June 2018

  @MasMarius
  Hi,

  Perhaps the Best Practices will help you.

  Here I am getting confused, should i go straight after picard step to HaplotypeCaller in ERC GVCF mode and then use GenomicDBImport to pass the GenotypeGVCFs for joint calling?

  You should run BQSR before variant calling. The slide decks in the Presentations section will help with the workflow.

  -Sheila

  1 · Share on Twitter Share on Google+
• MasMarius Member

  July 2018

  @Sheila
  Going to try it, really appreciate it!

  0 · Share on Twitter Share on Google+