OPIG members

Charlotte Deane

I lead the Oxford Protein Informatics Group (OPIG), a research group of over 40 people working on diverse problems across immunoinformatics, protein structure and small molecule drug discovery; using statistics, AI and computation to generate biological and medical insight.

My research covers several areas in protein structure prediction and protein interaction networks, combining both theoretical work and empirical analyses. We work on developing novel methodologies to understand and predict protein evolution, interaction, structure and function. Our work is focussed in the three main areas: protein structure, immunoinformatics and small molecules.

In OPIG, we develop novel algorithms, tools and databases that are openly available to the community. These tools are widely used web resources and are also part of several Pharma drug discovery pipelines.

Garrett Morris

I am particularly interested in bridging the fields of chemistry and AI/ML through better theory, model architectures, and novel representations. My research focuses on small-molecule structure-based drug discovery, especially:

• Docking and co-folding
• Virtual screening
• Target prediction
• De novo design
• Machine learning interatomic potentials (MLIPs)
• ADME-Tox and molecular property prediction
• Novel cheminformatics representations
• Novel data-splitting techniques
• The impact of noise and uncertainty quantification
• Explainable AI
• Applications of large language models (LLMs), including foundational, chemical, and genomic language models
• Agentic AI

Fergus Imrie

My research focuses on developing machine learning methods and techniques for medicine and drug discovery. I am particularly interested in approaches for designing potent, selective small molecules using structure-based methods, experimental design and decision-making in drug discovery, and how to learn efficiently from small quantities of (typically) noisy data.

Isak Valsson

My research interests involve developing robust machine learning methods for early drug-discovery problems. Recently, I’ve been working on developing structure based scoring functions that work better in an out-of-distribution (OOD) setting, i.e. when the training data occupies a different area in chemical space than the test data. Additionally, I’ve been exploring different ways to benchmark the OOD performance of binding affinity predictors, and different ways of featurising bound protein-ligand complexes.

Kate Fieseler

A fragment screen offers an information rich starting point for derivative compounds that can recapitulate fragment interactions. I am interested in how to maximally explore the fragment merge-design space with in silico design. By prioritizing synthetic tractability, fragment derivate designs can be synthesized via high-throughput multi-step chemistry reducing the overall cost and time to experimentally test them. In collaboration with XChem at Diamond Light source, I work directly with organic chemists and structural biologists to drive their fragment development projects forward.

Sam Money-Kyrle

Accurate prediction of molecular properties, such as protein-ligand affinity, off-target binding, toxicity, and mutagenicity, is an intrinsic component of small molecule drug discovery. My research focuses on the development of robust and generalisable computational tools that elucidate greater structural understanding of the interactions between small molecules and proteins. I am particularly interested in the application of novel deep learning methods and evaluating whether these approaches are capable of learning inherent biochemistry binding interactions.

Alexander Hasson

The Eric O’Neill lab recently identified a combination of biologically-targeted agents, that could restore epigenetic control and revert squamous pancreatic cancer cells to a more regulated differentiated state. Although we observe a shift towards normal cell epigenetics, this drug combination is not fully optimised to drive the phenotypic shift. In my research, I am developing and aim to use an innovative data-driven artificial intelligence (AI) approach to delineate novel molecules that can demonstrate improved and robust re-normalisation of epigenetic status and pancreatic differentiation, in order to offer new treatments for this hitherto intractable cancer. A major part of this work is the development of an inverse screening protocol, in order to find the protein targets of small molecules.

Arun Raja

My DPhil research involves the development of geometric deep learning methods for small molecule drug discovery grounded in physics and chemistry. Specifically, I am using deep learning for quantum-level representations of molecules as a precursor for tasks in lead molecule optimization such as property prediction and protein-ligand binding affinity prediction.

Charlie Clark

The chemokine signalling network drives inflammation and therefore many inflammatory diseases. However, it has evolved to be highly redundant to resist pathogenic shutdown, and so successful anti-chemokine therapeutics must target multiple chemokines simultaneously. During my DPhil, I will apply experimental and computational techniques to this ‘poly-pharmacological’ problem by characterising promiscuous therapeutics that can target multiple chemokines simultaneously and tackle chemokine-driven inflammatory disease.

Yael Ziv

My research focuses on AI methods for drug discovery. I work on diffusion and flow-matching models for structure-based drug design, exploring how protein structural information and design constraints can be incorporated into generative models to produce chemically and physically plausible molecules. I also work on agentic AI frameworks for therapeutic antibody optimization, exploring how AI agents can integrate computational tools for developability, humanization, and other design objectives into automated decision-making workflows.

Adelaide Punt

My research examines the impact of noise on small molecule activity prediction, identifying robust pairings of model and molecular embedding under increased artificial noise. By clustering molecules into chemical domains, I introduce domain-specific noise to mimic real-world variability. Within a federated learning framework, I address challenges from heterogeneous data distributions and client-specific noise variability by applying noise mitigation methods across both pre-processing and training steps, aiming to remove and smooth experimental noise

Nicholas Runcie

I am a DPhil student working on large language models (LLMs) for chemistry. My research focuses on evaluating the capabilities of LLMs in chemistry and developing agentic systems for autonomous scientific discovery.

Sanaz Kazeminia

I focus on structure-based drug design (SBDD) which uses 3D protein structures to design small molecules in a pocket-aware manner. My interest is primarily on the application of diffusion models in this space, which gradually add noise from molecular structures in three-dimensional space to generate novel compounds that match target protein binding sites. The goal is to generate molecules with high binding affinities to their targets which are synthetically accessible and chemically sound. I am interested in exploring the application of constraints on these models, new architectures for complex protein-ligand systems and the integration of a large language model to make these tools accessible to medicinal chemists.

Alvaro Prat

I am a first year PhD student working in both Computational Statistics and Machine Learning (CSML) and Oxford Protein Informatics (OPIG) groups. My current research interests pivot around developing robust and scalable generative models for structured data. I am also interested in low data regimes and advanced learning frameworks which focus on maximising information retrieval and signal propagation. I am excited to work alongside my supervisors, Yee Whye Teh (Oxford & DeepMind), Garrett Morris & Charlotte Deane (Oxford) to develop cutting edge solutions to enable accelerated & reliable drug discovery.

James Broster

I develop machine learning models to improve molecular docking by generating biologically meaningful and physically plausible poses that recover key molecular interactions. My work focuses on creating tools for real-world drug discovery applications, accounting for speed, accuracy, and the ability to account for protein flexibility.

Jochem Nelen

I am a Postdoctoral Researcher supporting the OpenBind consortium, working with Prof. Charlotte Deane and Dr. Fergus Imrie. OpenBind aims to generate large-scale experimental data on protein–ligand interactions to advance AI-driven drug discovery, and I contribute through computational methods and analysis. My research focuses on computational drug discovery, particularly virtual screening, molecular docking, and molecular dynamics to understand and predict protein–ligand interactions. I combine physics-based and machine learning approaches to improve the identification and prioritisation of active compounds.

Acer (Ace) Blake

I am a second-year DPhil student specializing in machine-learning for small-molecule drug discovery. My primary research interests focus on applications of machine learning for generative molecular design, molecular docking, and small-molecule representation learning.

I am particularly interested in the intersection between small-molecule drug discovery and biogerontology (the biology of ageing and age-related diseases). I am co-supervised by Professor Ghada Alsaleh within Oxford’s Nuffield Department of Orthopaedics, Rheumatology, and Musculoskeletal Sciences (NDORMS). Ghada’s lab work on the molecular biology of ageing, particularly with respect to the role of autophagy in age-related diseases.

Anita Love

Biological systems and processes are commonly reduced to pairwise graphs. This simplification often goes unquestioned or is assumed to have little impact on the task at hand. My research focuses on higher-order representations, such as hypergraphs, which capture these systems more faithfully, and on their potential to improve predictive power and contextual understanding.

I am a member of both OPIG and OSMANA, Gesine Reinert’s Oxford Stein’s Method and Network Analysis research group.

Kingsley Oguma

My work focuses on small-molecule drug discovery with an emphasis on molecular docking. I am particularly interested in the development and application of ranking algorithms for ranking protein-ligand binding affinities.

Matthew Raybould

My research applies immunoinformatics to improve therapeutic design and to better our understanding of the immune response. During my DPhil, I captured and compared structural representations of therapeutic and natural antibodies, leading to new structure-aware approaches for in silico developability assessment and screening library design. My research is now focused on incorporating structural awareness to improve our ability to identify broader sets antibodies with functional commonality and to define the functional boundaries of different classes of adaptive immune receptor.

Nele Quast

I develop and train deep learning models for T-cell receptor structures. I’m interested in training models that retain equivariance, merge sequence and structure information and can be injected with conditions or constraints. I’m also interested in the structure of the interface between TCRs and their pMHC antigen. Beyond my research I’m passionate about improving gender representation in STEM and have acted as president of the Oxford Wom*n in CS society.

Benjamin McMaster

T cells are a key part of our immune system, responsible for fighting pathogens and regulating immune responses. To identify foreign invaders, T cells use their T cell receptors (TCRs) to screen for and bind to antigens rapidly. Although key to our health and survival, the mapping between TCR composition and antigens remains poorly understood. I aim to apply newly developed deep learning models for protein structure prediction to TCR data to understand better the rules that govern antigen-specific T cell responses.

Henriette Capel

I am a final year DPhil student working on antibodies. My research is focused on leveraging computational tools to improve efficiency of antibody developability workflows. This includes the curation and maintenance of standardised datasets, and the generation of computational tools to guide antibody therapeutic development

Isaac Ellmen

Antibodies work by binding to their targets (antigens), and either inhibiting their function or activating other components of the immune system. Predicting the mode by which an antibody binds to its cognate antigen is called antibody-antigen complex modelling or docking. While general protein complex prediction has seen great improvements in recent years, driven by methods such as AlphaFold Multimer, antibody-antigen complexes are still difficult to model because we rarely have useful homologs to provide co-evolutionary information. My DPhil project is focused on developing new machine learning docking models to more accurately predict antibody-antigen complexes.

King Ifashe

I am a second year DPhil student, and my research sits at the intersection of molecular dynamics, machine learning, and structural immunology. I focus on understanding how T cell receptors recognise their targets, combining MD simulations and ML to study conformational behaviour and developing tools to predict binding and cross-reactivity, and more recently the design of novel protein sequences and structures. My work aims to have broad implications for immunotherapy and our fundamental understanding of immune recognition.

Marius Urbonas

Large deep-learning foundation models have become the dominant paradigm across many fields. Their ability to extract useful representations allows for efficient finetuning to target tasks. Even more interestingly we can predict exactly how the performance of these models will improve as we increase the model size or the amount of data without even training them, which motivated the building of ever larger language and vision models. I am interested in whether similar scaling laws can be found for immunology task modelling. My work aims to establish the data requirements for reliable immune modelling and guide more efficient therapeutic antibody discovery.

Odysseas Vavourakis

develop geometric generative models for protein structure, sequence, and dynamics, with a focus on de novo antibody design. The challenge is to generate human-like antibodies that specifically bind to a target while meeting a range of ancillary constraints—many related to antibody flexibility—in order to create viable therapeutics. I am therefore particularly focused on developing machine learning models to accurately predict conformational ensembles and enable robust antigen-conditional design. An interesting recurring challenge in the antibody space is the comparative scarcity of data compared to general-protein systems.

Clare Gillis

My research focuses on VHHs, also known as nanobodies – small immune proteins that function similarly to antibodies, but at roughly a tenth of the size and with half as many binding loops. Found naturally only in camelid species, nanobodies offer several advantages over conventional antibodies as therapeutics, including improved stability, better tumour penetration, the ability to cross the blood–brain barrier, and ease of manufacturing. I aim to better understand the properties of nanobodies, particularly what enables them to fold and function in the absence of a light chain, and to use this knowledge to expand the human nanobody design space.

Eric Wang

I am on the Immunoinformatics side. I am a first year, joint between Oxford and Scripps Research. I am working on the structural determinants of antibody neutralization of infectious diseases.

Lorenzo Tarricone

Symmetric protein assemblies, from vaccine nanoparticles to multi-subunit enzymes, are among the most functionally important and therapeutically promising structures in biology. Yet their computational design remains an unsolved problem, largely due to the complexity and scale of these systems. As a first year DPhil student, I am working at the intersection of deep generative modelling and structural biology, with the goal of making the de-novo design of large symmetric protein assemblies reliable and accessible. The long-term aim is a computational pipeline that can be meaningfully deployed in drug design and vaccine development.

Rebonto Haque

I am a second-year DPhil student working on pan-reactive antibody design. My work focuses on designing and validating anti-influenza antibodies, specifically looking at design principles to ensure designed antibodies are resistant to the mutational drift of viral epitopes. I have previously worked on using sparse autoencoders for the mechanistic interpretability of antibody language models.

Gabriel Abrahams

Proteins are remarkable nano-machines that carry out the myriad functions required for life to exist. Engineering proteins to have novel functions has a vast range of applications, ranging from medical developments such as combating anti-microbial resistance, to climate friendly industrial manufacturing. In my DPhil, I am working to develop a machine learning pipeline to steer directed evolution: a method for utilising the power of natural evolution to produce proteins with desirable capabilities that were not required to survive in nature. These experiments will be performed in the lab, in a massively high throughput screening platform currently being developed by the Engineered Biotechnology Research Group at Oxford.

Alexi (Hussain) Siddiqui

Understanding protein function requires the probing of both structure and dynamics. Traditional methods have limitations when attempting to capture dynamic behaviour. Hydrogen-Deuterium Exchange Mass-Spectrometry (HDX-MS) quantitatively assesses conformational dynamics and empirical models have been used to link the data to molecular dynamics (MD) simulations, potentially offering a more complete view of protein behaviour. My research is focused on reliable and robust methods for generating accurate conformations relevant with respect to experimental HDX-MS data. This is relevant to recently released structural prediction models. Using the wealth of existing data, we want to apply these methods for new insights. All developed code will be released, readily adaptable to existing analysis pipelines.

Qurat al ain (Annie)

Current approaches for protein design require multiple iterations of the design-make-test experimental cycle and provide limited control over the properties of the resulting molecules. I’m developing deep learning-based methods for designing de novo proteins with specific physicochemical properties. Computationally, this becomes a multi-objective optimisation problem where the output must be novel, diverse and physically plausible - how exciting! Feel free to reach out if you’re interested in similar topics and would like to have a chat.

Aaron Maiwald

My research focuses on understanding how neural networks process and learn from genomic data. Genomic language models such as the Nucleotide Transformer have been pretrained on vast amounts of DNA data and show promising performance on a range of benchmarks. As these systems become increasingly powerful and widely used in biological research, it’s crucial to understand exactly how they arrive at their predictions. Using techniques from mechanistic interpretability - an emerging field that reverse-engineers neural networks - I develop methods to reveal how these systems represent and transform biological information. This work is aimed to help us build more reliable and capable AI systems in genomics, and biology more broadly. If any of this interests you, please feel invited to reach out!

Hew Phipps

My DPhil research explores generative Flow models for protein folding making use of Molecular Dynamics (MD) simulations as training data. I studied Biochemistry at the University of Sheffield receiving an integrated Masters (MSci) with distinction in 2021 before moving to industry where I led the development of machine learning models for sequence-based protein directed evolution in Cambridge.

Ben Williams

I am a Research Software Engineer within OPIG where I work on developing, deploying and maintaining the group’s computational tools and resources, including the Structural Antibody Database (SAbDab). I hold a doctorate in condensed matter physics from Oxford and previously worked at Diamond Light Source, developing scientific software and data-analysis methods.

Christopher Taylor

I am one of OPIG’s Research Software Engineers (RSEs), assisting with the development, deployment, and maintenance of the software tools created within the group, as well as providing support to the community and our collaborators.  While this activity spans the breadth of the OPIG research portfolio, my own research interests are primarily in small molecule computational chemistry as applied to property prediction and molecular design.  I completed my PhD at the University of Bristol in 2014, and between 2014 and 2025 I was a postdoctoral research fellow at the University of Southampton.