microsoft / Elevation

End-to-end guide design for CRISPR/Cas9 with machine learning
MIT License
126 stars 34 forks source link

Scoring mismatches directly #4

Open augustboyle opened 5 years ago

augustboyle commented 5 years ago

Hello,

Thanks for making this code available. I am interested in the predicted activity of various combinations of mismatches on Cas9 activity on a synthetic DNA library. In short I want to give an sgRNA sequence and a DNA off-target and predict the activity. I have installed most of the package, which is fairly time intensive counting all the downloads and dependencies, but now the 'Predict' function won't run without the 3rd data dependency for the indexed human reference genome, which seems like another rather time intensive step.

Is there a way to score the sgRNA/DNA directly? This should be independent of the human genome and in fact I do not want any metrics from the genome to contaminate the prediction.

Evan