Research
Peter W. Higgs was a physical mathematician famous for using mathematical theories to predict the existence of a hitherto unknown particle called the Higgs particle, and popularly referred to as the god particle.
The motivating question in my research career has been:
βCan we cure disease and reverse-engineer cell and tissue perturbations using mathematically designed AI, just like Higgs used math to predict new particles?β
I am interested in causal inference, generative AI, disentangled representation learning, and applications of machine learning in biomedicine. Previously I studied string theory, quantum gravity and experimental supersymmetry. (Pure math papers use alphabetical author ordering.)
For an overview of my research trajectory you can check my talk at Yale Univesity,
Inference of Nature across disciplines.
|
|
Celcomen: spatial causal disentanglement for single-cell and tissue perturbation modeling
Stathis Megas, Daniel G. Chen, Krzysztof Polanski, Moshe Eliasof, Carola-Bibiane Schonlieb, Sarah A. Teichmann
arXiv, Machine Learning (cs.LG), 2024
|
|
Molecular connectomics: Placing cells into morphological tissue context
Stathis Megas, Nadav Yayon, Kerstin B. Meyer, Sarah A. Teichmann
PLOS Biology, 2024
|
|
EmptyDropsMultiome discriminates real cells from background in single-cell multiomics assays
Stathis Megas, Valentina Lorenzi, John C. Marioni
Genome Biology, 2024
|
|
2d TQFTs and baby universes
John Gardiner, Stathis Megas
Journal of High Energy Physics, 2021 (52)
|
|
Anomalous dimensions from thermal AdS partition functions
Per Kraus, Stathis Megas, Allic Sivaramakrishnan
Journal of High Energy Physics, 2020 (149)
|
|
Calculation of the single lepton SUSY analysis limits in the cMSSM m0-m1/2 plane for the CMS expriment at CERN
Stathis Megas
CMS twiki, 2014
|
Talks
I am passionate about giving talks about my work and making scientific knowledge broadly accessbile.
|
|