Debora Marks

Timeline

1. 1980

   Medical studies at Bristol

   She leaves medical study at the University of Bristol after her second MBChB degree in 1980.

2. 1993

   Mathematics at Manchester

   After returning to academia, she earns a BSc with honours in mathematics at Manchester University in 1993.

3. By 2010

   Harvard bioinformatics and a PhD

   After the Human Genome Project, she takes a bioinformatics position at Harvard. Her work on microRNA targeting becomes a PhD thesis submitted at Humboldt University in Berlin in 2010; during the PhD, she co-discovers the first microRNA in a virus.

4. 2011–2014

   Protein structures from sequence

   Between 2011 and 2012, she coauthors papers that use maximum-entropy models of evolutionary sequence variation to infer residue contacts and predict protein structures from sequence, including 11 transmembrane proteins. In 2014, she coauthors an eLife study that uses sequence coevolution to identify contacts and predict protein-complex structures.

5. 2016–2017

   Overton Prize and EVmutation

   In 2016, she receives the ISCB Overton Prize while serving as an assistant professor and director of the Raymond and Beverly Sackler Laboratory for Computational Biology at Harvard Medical School. In 2017, she coauthors EVmutation, an unsupervised statistical method that predicts mutation effects by modeling dependencies between residues.

6. 2021

   EVE for disease variants

   In 2021, she coauthors EVE, a deep generative model that predicts variant pathogenicity without disease labels; the paper estimates the pathogenicity of more than 36 million variants across 3,219 disease genes.

7. 2022

   Tranception for protein fitness

   In 2022, she coauthors Tranception, an autoregressive transformer that retrieves homologous sequences at inference to predict protein fitness.

8. 2025

   PopEVE and rare disease genetics

   In 2025, she coauthors popEVE, a deep generative model that combines evolutionary and human population data to estimate variant deleteriousness across the proteome. The model can prioritize likely causal variants using child exomes, even without parental sequencing.

9. November 2025

   A Pritzker Prize for AI

   In November 2025, she receives the inaugural Margot and Tom Pritzker Prize for AI in Science Research Excellence for work applying language models to evolutionary data and predicting phenotypes and molecular properties.

10. As of October 2026

    A Kempner award and Royal Society

    In January 2026, she and Noor Youssef receive a Kempner Institute Accelerator Award for a project to build a next-generation protein language model for viral evolution. She joins the Royal Society's 2026 Fellowship cohort. As of October 2026, she is Professor of Systems Biology at Harvard Medical School, Associate Faculty at the Broad Institute of Harvard and MIT, and Affiliate Faculty at Harvard's Kempner Institute.