Category Archives: Cheminformatics

GPT-5 achieves state-of-the-art chemical intelligence

I have run ChemIQ (our chemical reasoning benchmark) on GPT-5. The model achieves state-of-the-art performance with substantial improvements in the ability to interpret SMILES strings. Read my analysis and initial findings below. Scroll to the end for some cool demos.

Figure 1: Success rates for each model on the ChemIQ reasoning benchmark. Horizontal brackets between adjacent bars indicate the result of a two-tailed McNemar’s test comparing paired outcomes for the same questions. Significance levels are shown as: n.s. (not significant, p ≥ 0.05), * (p < 0.05), ** (p < 0.01), and *** (p < 0.001).

Continue reading →

A more robust way to split data for protein-ligand tasks?

As I was recently reading through the paper on the PLINDER dataset while preparing for my next project, one of the aspects of the dataset that caught my attention was how the dataset splits were done to ensure minimal leakage for various protein-ligand tasks that PLINDER could be used for. They had task-specific splits as the notion of data leakage differed from task to task. For instance, in rigid body docking, having a similar protein in the train and test may not be considered leakage if the binding pocket location, conformation, or pocket interactions with a ligand are significantly different. On the other hand, in the case of co-folding, having similar proteins in the train and test sets would be considered data leakage, as predicted protein structures play a significant role in accuracy scoring. The effort that went into creating task-specific splits resonates strongly with OPIG’s view on ensuring minimal data leakage for validating the generalisability of protein-ligand models. However, it may become tedious to create task-specific dataset splits for every protein-ligand task when dealing with a large suite of such tasks. This had me thinking of potential avenues to streamline the dataset split process across the tasks, and one way to do this is by using protein-ligand interaction fingerprints or PLIFs.

Continue reading →

ChatGPT can now use RDKit!

All chemistry LLM enthusiasts were treated to a pleasant surprise on Friday when Greg Brockman tweeted that ChatGPT now has access to RDKit. I’ve spent a few hours playing with the updated models and I have summarized some of my findings in this blog.

Continue reading →

Geometric Deep Learning meets Forces & Equilibrium

Introduction

Graphs provide a powerful mathematical framework for modelling complex systems, from molecular structures to social networks. In many physical and geometric problems, nodes represent particles, and edges encode interactions, often acting like springs. This perspective aligns naturally with Geometric Deep Learning, where learning algorithms leverage graph structures to capture spatial and relational patterns.

Understanding energy functions and the forces derived from them is fundamental to modelling such systems. In physics and computational chemistry, harmonic potentials, which penalise deviations from equilibrium positions, are widely used to describe elastic networks, protein structures, and even diffusion processes. The Laplacian matrix plays a key role in these formulations, linking energy minimisation to force computations in a clean and computationally efficient way.

By formalising these interactions using matrix notation, we gain not only a compact representation but also a foundation for more advanced techniques such as Langevin dynamics, normal mode analysis, and graph-based neural networks for physical simulations.

Continue reading →

Narrowing the gap between machine learning scoring functions and free energy perturbation using augmented data

I’m delighted to report our collaboration (Ísak Valsson, Matthew Warren, Aniket Magarkar, Phil Biggin, & Charlotte Deane), on “Narrowing the gap between machine learning scoring functions and free energy perturbation using augmented data”, has been published in Nature’s Communications Chemistry (https://doi.org/10.1038/s42004-025-01428-y).


During his MSc dissertation project in the Department of Statistics, University of Oxford, OPIG member Ísak Valsson developed an attention-based GNN to predict protein-ligand binding affinity called “AEV-PLIG”. It featurizes a ligand’s atoms using Atomic Environment Vectors to describe the Protein-Ligand Interactions found in a 3D protein-ligand complex. AEV-PLIG is free and open source (BSD 3-Clause), available from GitHub at https://github.com/oxpig/AEV-PLIG, and forked at https://github.com/bigginlab/AEV-PLIG.

Continue reading →

Molecule Networks: data visualization using PyVis

Over the past few years I have explored different data visualization strategies with the goal of rapidly communicating information to medicinal chemists. I have recently fallen in love with “molecule networks” as an intuitive and interactive data visualization strategy. This blog gives a brief tutorial on how to start generating your own molecule networks.

Continue reading →

A tougher molecular data split – spectral split

Scaffold splits have been widely used in molecular machine learning which involves identifying chemical scaffolds in the data set and ensuring scaffolds present in the train and test sets do not overlap. However, two very similar molecules can have differing scaffolds. In an example provided by Pat Walters in his article on splitting chemical data last month, he provides an example where two molecules just differ by a single atom and thus have a very high Tanimoto similarity score of 0.66. However, they have different scaffolds (figure below).

In this case, if one of the molecules were in the train set and the other in the test set, predicting the test molecule would be quite trivial as there is data leakage. Therefore, we need a better splitting method such that there is minimal overlap between the train and test set. In this blogpost, I will be discussing spectral split, a splitting method introduced by our fellow OPIG member, Klarner et. al (2023).

Spectral split

Spectral split or clustering is based on the spectral graph partitioning algorithm. The basic idea of spectral clustering is as follows: The dataset is projected on a R^n matrix. An affinity matrix using a kernel that could be domain-specific is defined. Following that, the graph Laplacian is computed from the affinity matrix, followed by its eigendecomposition. Then,  k eigenvectors corresponding to the k lowest/highest eigenvalues are selected. Finally, the clusters are formed using k-means.

In the context of molecular data splitting, one could use the Tanimoto similarity metric to construct a similarity matrix between all the molecules in the dataset. Then, a spectral clustering method could be used to partition the similarity matrix such that the similarity within the cluster is maximized whereas the similarity between the clusters is minimized. Spectral split showed the least overlap between train (blue) and test (red) set molecules compared to scaffold splits (figure from Klarner at. al. (2024) below)

In addition to spectral splits, one could attempt other tougher splits one could attempt such as UMAP splits suggested by Guo et. al. (2024). For a detailed comparison between UMAP splits and other commonly used splits please refer to Pat Walters’ article on splitting chemical data.

Controlling PyMol from afar

Do you keep downloading .pdb and .sdf files and loading them into PyMol repeatedly?

If yes, then PyMol remote might be just for you. With PyMol remote, you can control a PyMol session running on your laptop from any other machine. For example, from a Jupyter Notebook running on your HPC cluster.

Continue reading →

The XChem trove of protein–small-molecules structures not in the PDB

The XChem facility at Diamond Light Source is truly impressive feat of automation in fragment-based drug discovery, where visitors comes clutching a styrofoam ice box teeming with apo-form protein crystals, which the shifter soaks with compounds from one or more fragment libraries and a robot at the i04-1 beamline kindly processes each of the thousands of crystal-laden pins, while the visitor enjoys the excellent food in the Diamond canteen (R22). I would especially recommend the jambalaya. Following data collection, the magic of data processing happens: the PanDDA method is used to find partial occupancy in the density, which is processed semi-automatedly and most open targets are uploaded in the Fragalysis web app allowing the ligand binding to be studied and further compounds elaborated. This collection of targets bound to hundreds of small molecules is a true treasure trove of data as many have yet to be deposited in the PDB, making it a perfect test set for algorithm design: fragments are notorious fickle to model and deep learning models cannot cheat by remembering these from the protein database.

Continue reading →

Out of the box RDKit-valid is an imperfect metric: a review of the KekulizeException and nitrogen protonation to correct this

In deep learning based compound generation models the metric of fraction of RDKit-valid compounds is ubiquitous, but is problematic from the cheminformatics viewpoint as a large fraction may be driven by pyrrolic nitrogens (see below) rather than Texas carbons (carbon with 5 bonds like the Star of Texas). In RDKit, no error is more irksome that the KekulizeException or ValenceException from RDKit sanitisation. These are raised when the molecule is not correct. This would make the RDKit-valid a good metric, except for a small detail: the validity is as interpreted from the the stated implicit and explicit hydrogens and formal charges on the atoms, which most models do not assign. Therefore, a compound may not be RDKit-valid because it is actually impossible, like a Texas carbon, but in many cases it is because the formal charge or implicit hydrogen numbers of some atoms are incorrect. In both case, the major culprit is nitrogen. Herein I go through what they are and how to fix them, with a focus on aromatic nitrogens.

Continue reading →