15/03/2024
I built a machine-learning model as well as a random forest to evaluate CHEMBL's data for drug pe*******on into the bacteria. I was surprised to see that LogP (a measure of the lipophilicity of a drug) was given more weightage by the model over Lipinski's rule of 5. One second thought it makes sense because the model already accounts for the components of Lipinski's rules and tends to minimize its contribution. My model was 88 percent accurate in predicting a drug's ability to enter bacteria.
