Runtime on CCS reads

[Lionward]

Do you have an estimate on the runtime on CCS reads based on the current version?

In your study you mentioned that It's suppose to be 3 times faster than the standard whatsapp-Deepvariant approach.

In this study: https://github.com/google/deepvariant/blob/r1.6.1/docs/metrics.md. it's shown that Deep variant requires 6.65h on CCS reads using 64 cpus, would then Pepper be able to complete the analysis in ~2,2h?

Best!

[Lionward]

Benchmark results were found here:
https://static-content.springer.com/esm/art%3A10.1038%2Fs41592-021-01299-w/MediaObjects/41592_2021_1299_MOESM1_ESM.pdf

I am still wondering since for the r0.6 releaes, you mentioned that there was 2x acceleration on PacBio-HiFi variant calling pipeline. Does that mean that now it will take 1,1 h using 64 cores.

Best!