Home

Research Interests

I am a Senior Machine Learning Scientist in Roche’s AI for Drug Discovery team in Basel. I work at the intersection of deep learning and structural biology, developing methods for de novo antibody and protein design. My research centres on two main areas:

Mimicking natural interactions. I develop algorithms that design antibodies to mimic how other proteins bind their targets. The aim is to use these natural interactions to guide antibody design towards specific epitopes. This work includes Mimic antibodies and MIMOSA.

Scoring antibody–antigen complexes. Alongside design, I build models to score antibody–antigen complexes and evaluate predicted binding interactions. I also helped develop POISK, a dataset of antibody sequences and functional metadata extracted from patent literature that supports the evaluation of antibody–antigen complex prediction. You can browse the data using the POISK Explorer.

I completed my DPhil in Statistics at the University of Oxford, working with Prof. Charlotte Deane in the Oxford Protein Informatics Group (OPIG). My doctoral research focused on antibody structure prediction using equivariant deep learning. Key projects included ABlooper for predicting antibody CDR-loop structures and ImmuneBuilder, a suite of tools for predicting the structures of antibodies, nanobodies and T-cell receptors. You can try ImmuneBuilder through the group’s web server. I also contributed to KA-Search, a method for rapidly searching large antibody datasets.

Education

---

Other interests

I grew up in a tiny village in the north of Spain with three ugly brothers and two beautiful dogs. I like all types of kayaking, visiting new places, discovering new books and playing board games. Here are some photos!!!