Scout: prediction effect for in frame ins/del

[4WGH]

could this be implemented in SCout? See publication link
This would complement the severity prediciton when SIFT and polyphen can not

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0077940

[moonso]

Think that this is a question for @henrikstranneheim if it should be included in MIP

[moonso]

Please have a look at that @henrikstranneheim

[dnil]

It operates with the at first sight counterintuitive result that changes to repeat stretches are likely to affect function. Unless in an unordered part of the protein. Interesting!

[dnil]

Fig 3 reconciles the world views: one-aa changes to a e.g. a polyQ is not so likely to be disease-causing, but two much more so.

[4WGH]

question about prediction of in frame deletion come up again.
something to look into?

[dnil]

One version seen operational in Lund; Paul P has a working setup running CADD locally on indels - I'm sure he would help iron out any issues we may run into there. SIFT Indel looks nice enough, but CADD seems like a great place to start.

[henrikstranneheim]

IN MIP decided to:

• Skip SIFT locally for indels as it is not practical to implement in a production setting and will most likely add less than CADD.
• Added dynamic calculation of CADD for indels. VCF key "CADD" as usual.

[dnil]

Ok, if the keys are as for regular SNVs, we simply close this issue! Thank you - this should be helpful! :+1: