Zero-shot mutation effect prediction on protein stability using RFdiffusion (Similar to RFjoint in RFpaper)

Hi RFdiffusion team,

I came across the paper "Zero-shot mutation effect prediction on protein stability and function using RoseTTAFold", where RFjoint was used for zero-shot mutation effect prediction.

I was wondering:

1. Is it possible to perform zero-shot mutation effect prediction on protein stability using RFdiffusion?
2. If not currently supported, would it be feasible to extend RFdiffusion for such a task?
3. Do you have any recommendations on how one might adapt RFdiffusion for this purpose?
4. Would you suggest using RFjoint instead for such analyses, or is there a better alternative?

I appreciate any insights you can provide on this! Thanks for your amazing work on RFdiffusion.

Best regards,
Ali

The issue names no files, tests, or entry points and asks whether RFdiffusion can support zero-shot mutation effect prediction. Start by reviewing the cited RFjoint and RoseTTAFold work alongside the repository's existing prediction workflows. A concrete implementation target and maintainer guidance would be needed before defining what done means.