search

Gleeson-Lab / wxs_pipeline

Starting with BAMs and FASTQs, follow GATK 4.0 Best Practices up to generating a joint-genotyped VCF
1 stars 1 forks source link

Variant Annotation #13

Open brcopeland opened 3 years ago

brcopeland commented 3 years ago

I have the most previous experience with SnpEff/ClinEff. Do people here have preferences as to which tool if any to run and any non-standard parameters to use?

shishenyxx commented 3 years ago

You can re-implement this pipeline used for 4dbsm and sperm project. This basically contains all annotations we used for the mosaic-related projects (only for variant filtering).

For functional annotations, I normally use ANNOVAR, the command would be something like /home/xiy010/tools/annovar/table_annovar.pl /projects/ps-gleesonlab5/user/xiy010/analysis/2018-12-11-Amplicon/2019-09-13-new_full/1-Alignment_and_germline_calling/FCD_germline_variant/2019_06_10_Ampliseq_full.vcf /home/xiy010/tools/annovar/humandb/ -buildver hg19 -out 2019_06_10_Ampliseq_annotated -remove -protocol refGene,gnomad_genome,avsnp150,cosmic89,cadd13,eigen,fathmm -operation g,f,f,f,f,f,f -nastring . -vcfinput and annotate as many tables as possible ... You can take a look at the databases used for Varvis or SEQR ... and download them .... btw I downloaded most of the hg19 vers in the folder indicated above already ... You can also ask Sangmoon what he is using for annotations.

The problem for annotation is you sometimes have to switch between different versions of the databases, for example, dbnsfp after a certain version (v3.0 or so) replaces SIFT and Polyphen2 with PROVEAN ... but some reviewers will still ask for SIFT and Polyphen ... so we either fix a list of database that we use (like the avsnp150, cadd13, and COSMIC89 and so on). Also, have to download everything separately for GRCh38 ...

brcopeland commented 3 years ago

As you say the desired annotation depends on the circumstances. If we can agree on a reasonable default I can add that. If it is instead too dependent on the individual circumstances then perhaps we should omit this.

shishenyxx commented 3 years ago

Yeah depends on whether we want an alternative for Varvis/SEQR ... etc