Joe Marsh

How genetic variation in protein-coding genes leads to human disease.

Joe Marsh is Professor of Computational Protein Biology at the University of Edinburgh. He leads a research group focused on understanding how genetic variation perturbs protein structure, interactions, and function. 

His work sits at the interface of protein biophysics, human genetics, and computational biology, with a particular emphasis on interpreting disease-associated mutations.

photo of Joe Marsh
Joe Marsh

Dr Marcin Plech, Postdoctoral Research Fellow (joint with Grzegorz Kudla)
Dr Benjamin Livesey, Postdoctoral Research Fellow
Dr Lukas Gerasimavicius, Postdoctoral Research Fellow
Dr Rolando Hernandez Trapero, Postdoctoral Research Fellow
Dr Hasan Çubuk, Postdoctoral Research Fellow
Ankit Pathak, PhD student
Verena Obermüller, PhD student (joint with Hannah Long)
Rowena Tao, PhD student
Yifei Shang, PhD student
Tesni Walsh, PhD student
Amber Minhas, PhD student (joint with Georg Kustatscher)
 


The Marsh lab studies how genetic variation in protein-coding genes leads to human disease. 

The work spans three complementary areas: evaluating, benchmarking, and improving the interpretation and clinical utility of variant effect predictors; understanding and predicting the diverse molecular mechanisms by which disease-causing variants act, including loss-of-function, gain-of-function, dominant-negative, and hypomorphic effects; and using multiplexed assays of variant effect to systematically measure the functional consequences of large numbers of mutations. 

Together, these efforts aim to move variant interpretation beyond categorical classification toward a quantitative, mechanism-centred understanding of how sequence variation perturbs protein function.


structure of proteins binding DNA

Gerasimavicius L, Biddie SC & Marsh JA (2026) A structure-guided approach to non-coding variant evaluation for transcription factor binding using AlphaFold 3. Nucleic Acids Research 10.1093/nar/gkaf1417

Badonyi M & Marsh JA (2025) Prevalence of loss-of-function, gain-of-function and dominant-negative mechanisms across genetic disease phenotypes. Nature Communications 10.1038/s41467-025-63234-3

Livesey BJ & Marsh JA (2025) Variant effect predictor correlation with functional assays is reflective of clinical classification performance. Genome Biology26:104 10.1186/s13059-025-03575-w

Chillón-Pino D, Badonyi M, Semple CA & Marsh JA (2024) Protein structural context of cancer mutations reveals molecular mechanisms and candidate drivers. Cell Reports 43:114905


This article was published on