# Meaningful machine learning models and machine-learned pharmacophores from fragment screening campaigns

arXiv: https://arxiv.org/abs/2204.06348 · HTML: https://mlsanity.com/paper/2204.06348/

## Key facts

- No citing papers in the map yet (data through 13 Oct 2026).
- Builds on 8 of its 36 references that are in the map.
- First posted on arXiv 25 Mar 2022.

## Builds on (most cited first, 8 of 8)

- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [Automatic chemical design using a data-driven continuous representation of molecules](https://mlsanity.com/paper/1610.02415/) (2016)
- [Molecular Graph Convolutions: Moving Beyond Fingerprints](https://mlsanity.com/paper/1603.00856/) (2016)
- [AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery](https://mlsanity.com/paper/1510.02855/) (2015)
- [Using Attribution to Decode Dataset Bias in Neural Network Models for Chemistry](https://mlsanity.com/paper/1811.11310/) (2018)
- [Autonomous discovery in the chemical sciences part I: Progress](https://mlsanity.com/paper/2003.13754/) (2020)
- [Noisy, sparse, nonlinear: Navigating the Bermuda Triangle of physical inference with deep filtering](https://mlsanity.com/paper/1912.04345/) (2019)
- [BenchML: an extensible pipelining framework for benchmarking representations of materials and molecules at scale](https://mlsanity.com/paper/2112.02287/) (2021)

## Abstract (arXiv)

Machine learning (ML) is widely used in drug discovery to train models that predict protein-ligand binding. These models are of great value to medicinal chemists, in particular if they provide case-specific insight into the physical interactions that drive the binding process. In this study we derive ML models from over 50 fragment-screening campaigns to introduce two important elements that we believe are absent in most -- if not all -- ML studies of this type reported to date: First, alongside the observed hits we use to train our models, we incorporate true misses and show that these experimentally validated negative data are of significant importance to the quality of the derived models. Second, we provide a physically interpretable and verifiable representation of what the ML model considers important for successful binding. This representation is derived from a straightforward attribution procedure that explains the prediction in terms of the (inter-)action of chemical environments. Critically, we validate the attribution outcome on a large scale against prior annotations made independently by expert molecular modellers. We find good agreement between the key molecular substructures proposed by the ML model and those assigned manually, even when the model's performance in discriminating hits from misses is far from perfect. By projecting the attribution onto predefined interaction prototypes (pharmacophores), we show that ML allows us to formulate simple rules for what drives fragment binding against a target automatically from screening data.