Tumor-normal mode with rnaseq data

[gianfilippo]

Hi,

is ti possible to run deepsomatic on matched tumor-normal rnaseq samples ? from a previous issue, it seems that the WES mode is the way to run this case. Is this tested ? In deepvariant the "customized_model" options is used to point to a a model trained using RNA-seq data. Could you please clarify ?

Thanks

[kishwarshafin]

Hi @gianfilippo ,

The best suited model for your use-case right now is the WES tumor-normal model. You can find the case-study here: https://github.com/google/deepsomatic/blob/r1.7/docs/deepsomatic-case-study-wes.md.

We have not tested our models on rnaSEQ data as we didn't have benchmarking data. The DeepVariant RNASEQ model is for germline variants and is not suitable for somatic variant calling.

[gianfilippo]

thanks. The lack of rnaseq benchmarking data is definitely a problem

On Wed, Sep 18, 2024 at 11:09 AM Kishwar Shafin @.***> wrote:

Hi @gianfilippo https://github.com/gianfilippo ,

The best suited model for your use-case right now is the WES tumor-normal model. You can find the case-study here: https://github.com/google/deepsomatic/blob/r1.7/docs/deepsomatic-case-study-wes.md .

We have not tested our models on rnaSEQ data as we didn't have benchmarking data. The DeepVariant RNASEQ model https://github.com/google/deepvariant/blob/r1.6.1/docs/deepvariant-rnaseq-case-study.md is for germline variants and is not suitable for somatic variant calling.

— Reply to this email directly, view it on GitHub https://github.com/google/deepsomatic/issues/23#issuecomment-2358739612, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACPSFVGH2XPQVZJYMVAH5ELZXGJRLAVCNFSM6AAAAABOIIEZ5OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNJYG4ZTSNRRGI . You are receiving this because you were mentioned.Message ID: @.***>