How does DeepMosaic perform on PCR-amplified WES data?

[sattarkh]

DeepMosaic seems like a very powerful tool. I may have missed it, but it looks in the paper you have trained the model on PCR-amplified libraries and tested it on PCR-free libraries. How does DeepMoasic perform if tested on PCR-amplified libraries, specifically WES data?

[shishenyxx]

Dear Dr. Sattar,

Thank you for your interests in DeepMosaic, the software is trained with PCR-amplified WGS libraries and computationally simulated data and tested on other PCR-free or PCR amplified WGS libraries, because of the biased depth from exome capture, the current version and parameters are only optimized for genome data. We are currently testing DeepMosaic on PCR-amplified captured WES data, will release the recommended parameters and sensitivity measurement once we finish our internal testing!

Best,

Xiaoxu

[shishenyxx]

Hi Dr. Sattar,

Just an update that per our assessment, DeepMosaic (the current genome version) doubled the validation rate compared with our previous exome mosaic calling pipelines. While this is still only 40%. Given the high false-positive rates calling mosaic variants from WES, we think that's already providing a much more convincing alternative compared with other methods we have tried.

Best,

Xiaoxu

[sattarkh]

Hi Xiaoxu,

Thank you for the update. May I ask, what pipelines are you camping it to? Also, is this in tumor-only or paired mode?

Thanks, Sattar

On Wed, Nov 17, 2021 at 8:04 PM Xiaoxu Yang @.***> wrote:

Hi Dr. Sattar,

Just an update that per our assessment, DeepMosaic (the current genome version) doubled the validation rate compared with our previous exome mosaic calling pipelines. While this is still only 40%. Given the high false-positive rates calling mosaic variants from WES, we think that's already providing a much more convincing alternative compared with other methods we have tried.

Best,

Xiaoxu

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Virginiaxu/DeepMosaic/issues/1#issuecomment-972391827, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACDECO3YJS2K4EA5VYRL7J3UMRGLBANCNFSM4VOVF7PA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[shishenyxx]

Hi Sattar,

1. We are comparing it against GATK Polidy 50 + filters (PMID 33781308);
2. Tumor only, the tested one was Mutect2 single mode with an in-house panel of normal+DeepMosaic.

Best,

Xiaoxu

[sattarkh]

Perfect, thanks for that information. I am in the process of validating some M2 variants and I may give DeepMosaic a try after I have a validated callset.

Thanks again, Sattar

On Wed, Nov 17, 2021 at 8:56 PM Xiaoxu Yang @.***> wrote:

Hi Sattar,

1. We are comparing it with GATK Polidy 50 + filters (PMID 33781308)
2. Tumor only, the tested one was on Mutect2 single mode with an in-house panel of normal.

Best,

Xiaoxu

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Virginiaxu/DeepMosaic/issues/1#issuecomment-972429904, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACDECO6JJQMFS2NY6WN7GHTUMRMOHANCNFSM4VOVF7PA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[shishenyxx]

Hi Sattar,

Yeah please try and we welcome all feedback from you. You can feed DeepMosaic with the raw M2 output (maybe select "PASS" to increase the specificity) and we can see from there.

Best,

Xiaoxu

[sattarkh]

Will do, thanks.

On Wed, Nov 17, 2021 at 9:05 PM Xiaoxu Yang @.***> wrote:

Hi Sattar,

Yeah please try and we welcome all feedback from you. You can feed DeepMosaic with the raw M2 output (maybe select "PASS" to increase the specificity) and we can see from there.

Best,

Xiaoxu

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Virginiaxu/DeepMosaic/issues/1#issuecomment-972433706, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACDECO77S5TKC7P3TZICMA3UMRNPBANCNFSM4VOVF7PA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.