Practical Cheminformatics

As seems to be the case these days, this post was motivated by a comment I saw in the blogosphere.  In one of the myriad discussions on applications of AI in drug discovery, someone wrote:  "I have yet to encounter a machine learning algorithm which predicts a true scaffold hop (say from Viagra to Cialis). From that standpoint, a tool like ROCS which looks at abstract but general features like shape and electrostatics is better than a lot of ML." The comment got me thinking about something that has bugged me for a long time.   What exactly constitutes a scaffold hop?  Should we consider Viagra to Cialis a scaffold hop? (hint, I don't think so, stick with me and I’ll explain)  What is a scaffold hop? Let’s start by taking a step back and looking at some of the classic work of Hans-Joachim Böhm and Martin Stahl.  In their 2004 paper , Böehm and Stahl highlighted three different three-dimensional approaches to scaffold hopping.   The scaffold r

This post was partially motivated by a recent post from Karl Leswing describing how to use the DeepChem package to do virtual screening on a large database.  As part of the tutorial, Karl used the HIV sample file that is part of the DeepChem distribution to build a model.  This model was then used to select compounds from the ZINC database .  The tutorial is nice and the methodology is explained in a manner that is easy to follow.  The problem is that the molecules selected by the model are not what I would consider "drug-like". The molecules reported in Karl's post have aryl sulfonic acid groups.  In fact, molecule D has 2 sulfonic acids AND a thiol.  To someone who has spent a bit of time looking through HTS data, these molecules scream false positive .  If an enterprising young computational chemist were to take this set of hits to an experienced medicinal chemistry colleague, my guess is that he/she would simply shake his/her head and say something like &