Thoughts on “Predicting Electron Paths” (Bradshaw et al.)

A couple of weeks ago, we came across a preprint [] tackling the problem of forward prediction using a stepwise approach. The main premise is that reactions can be described as a linear sequence of electron pushing events, which a neural model can be trained to select/predict. The authors aim to develop a method for Read More

Large-Scale Testing of MIT Property Prediction Model

Regina: In our conversation with member companies, we have discovered that all of you are interested in a model for property prediction.  Our model based on graph convolution network is available and ready for you to use.  However, before transitioning it to your organizations, we wanted to test it on a large and varied collection Read More

ML Tutorial is on-line

I uploaded the tutorial on basics of ML that is similar to the one given during the consortium meeting.  It may be useful for people in your company who didn’t attend the meeting at MIT,  but would like to learn the subject matter. Please feel free to post your questions here, and we will reply Read More

Thoughts on Recent Nature Paper

First, let me update you on our news: Our paper on “Junction Tree Variational Autoencode for Molecular Graph Generation” has been accepted to ICML.  This is a paper that looks at novel ways for lead optimization. We are currently working on the final version. If you have any comments/questions, please send our way. On Thursday, Read More

Welcome to MLPDS blog!

To make our collaboration successful, we would like to stay in touch with you. This blog will be our on-line forum for updating you about recent developments, discussing papers in chemistry/ML space, and hearing about your experiences with using our tools.  Feel free to propose topics you would like to have covered. While the MIT Read More