How can categorical models provide value in supporting compound prioritisation?
Introduction In early-stage drug discovery, medicinal chemists rely on predictive models to help guide which compounds to synthesise or test…
Summary
This paper demonstrates how the Surflex-PSIM method can help investigate hypotheses around protein function in cases where function cannot be determined by sequence similarity.
Surflex-PSIM is a surface-based protein similarity algorithm, which supports fully automatic binding site detection and is fast enough to screen comprehensive databases of protein binding sites.
In this study, binding site detection was validated on apo/holo cognate protein pairs, and used to screen a set of 8 proteins that had poorly characterized functions at the time of crystallization, but were later biochemically annotated. It showed excellent performance identifying binding site matches, and for a panel of 12 unannotated proteins, supported suggestion of likely functions.
More publications to browse
Test and Prove the Value of AI in Drug Discovery
Defining value is the best place to start Before diving into the specifics of testing AI’s value, the first step…
How to evaluate the performance of QSAR/QSPR classification models?
Introduction After training a classification model, we would like to evaluate its performance by using the trained model on an…
