I found some interesting toxicology datasets from the Tox21 challenge, and wanted to see if it was possible to build a toxicology predictor using a deep neural
Peeking into the chemical space using free tools
As covered before, chemical space is huge. So it could be nice if this multidimensional molecular space could be reduced and visualized to get an idea about
Molecular neural network models with RDKit and Keras in Python
Neural networks are interesting models underlying much of the newest AI applications and algorithms. Recent advances in training algorithms and GPU enabled code together with publicly available
Teaching Computers Molecular Creativity
Neural Networks are interesting algorithms, but sometimes also a bit spooky. In this blog post I explore the possibilities for teaching the neural networks to generate completely
RDKit UGM 2016
I’m looking forward for the first to attend the RDKit user group meeting from 26-28 October 2016 in Basel, Switzerland. RDKit is an open source chemoinformatics toolkit
How to solve problems with coordinate bonds in Rdkit
When I have been working with chemical databases and import of molecules I have encountered numerous problems with the way chemical structures are drawn. Most often the
Learn how to map a simple Ames mutagenicity model to molecular features using RDkit.
Toxic compounds are most often something that we try to avoid when designing novel pharmaceutical compounds, so it could be nice to get a prediction if a
Learn how to hack RDKit to handle peptides with pseudo atoms
The chemoinformatics package Rdkit has is strength with handling small organic molecules. These molecules are characterized by a large diversity in chemical structures. A description of the
Safer fitting through regularization
Last time a simple multiple linear regression (MLR) model was seriously overfitted to molecular solubility data. This time the concept of regularization will be tested. Recall that
Wash that gold: Modelling solubility with Molecular fingerprints….
Last blog entry the conversion between molecule and fingerprint was briefly touched upon. Now the fingerprints will be used as the basis for a simple attempt to