Re:Keras-based molecular autoencoder
by rgbombarelli on Tuesday October 25, 2016 @12:38AM
(#12785332)
We did something like that for a more continuous property, like logP, which is easy to predict with cheminformatics. We are working on metrics that reflect drug-likeness better.
[ Parent
| on HN ]
Interesting. If I'm not mistaken, logP is a type of property where it is easy to look up constituents' "contribution" for the purpose of predicting logP with substitutions/alterations of the example molecule.
Re:Keras-based molecular autoencoder
by rgbombarelli on Tuesday October 25, 2016 @07:22AM
(#12786932)
Exactly, we have a good understanding that logP is additive (as a matter of fact, one predicts it using group contributions). The AE noticed this and started adding halogens to already high logP molecules.
The interesting part is that it stops before going totally crazy and per-halogenating the molecule. This is probably because it has an intuition about how molecules look like and hasn't really seen that kind of substitution pattern.
Re:Keras-based molecular autoencoder
by Houshalter on Thursday October 27, 2016 @12:31AM
(#12802925)
Wait, what is logP? I assumed you meant log probability, since that is the standard error metric for text prediction tasks like this.
Re:Keras-based molecular autoencoder
by rgbombarelli on Monday October 31, 2016 @12:11PM
(#12838153)
It's the water-octanol partition coefficient, a basic molecular descriptor of the physiological distribution of a drug. It is one of the multiple targets one aims to optimize in a novel drug-like compound.