search

jozhang97 / MutateEverything

62 stars 5 forks source link

Reverse pred for single mutation #8

Closed daiheng-zhang closed 3 months ago

daiheng-zhang commented 3 months ago

Modify the reversed prediction for single mutation when calculating loss1.

jozhang97 commented 3 months ago

Hi Daiheng,

Thanks for catching this bug and sorry for the slow response.

The root cause of this issue is that the ddG values in the data files are negated. I have uploaded the fixed data files (follow the same data preparation steps again).

In the new files, ddg < 0 means that the mutation was stabilizing. Let me know if you find any issues.

Thanks again!