Eli N. Weinstein


I’m a postdoctoral research scientist working with David Blei (link) at Columbia University. I also consult part-time at Jura Bio (link). I received my PhD in Biophysics from Harvard University in May 2022, advised by Debora Marks (link) and Jeff Miller (link), as a Hertz Foundation Fellow (link). I received my A.B. from Harvard in 2016, concentrating in Chemistry and Physics, and working with Adam Cohen (link).

I develop statistical tools for molecular biology, working in the fields of Bayesian statistics, probabilistic machine learning, biophysics and genomics. My primary applied interests are in generative statistical methods for biological sequences. In particular, I develop techniques to learn from complex observational sequence data and predict novel sequences that can be synthesized in the laboratory; these methods have application in evolutionary biology, immunology, microbiology, virology, clinical genetics and many other areas of biology, biomedicine and biotechnology. My primary theoretical interests are in Bayesian methodology, especially issues related to model misspecification and causal inference. I am especially interested in formalizing, analyzing, and generalizing heuristic methods and imprecise ideas from computational biology.

Email: ew2760 [at] columbia.edu

Selected publications and preprints

Eli N. Weinstein*, Alan N. Amin*, Jonathan Frazer, Debora S. Marks (*Equal contribution). Non-identifiability and the blessings of misspecification in models of molecular fitness and phylogeny. Accepted at NeurIPS 2022. Oral presentation. preprint. talk.

Eli N. Weinstein, Jeffrey W. Miller. Bayesian data selection. Accepted with minor revisions at the Journal of Machine Learning Research. IBM Student Paper Award at NESS 2021. preprint. code. talk.

Eli N. Weinstein, Alan N. Amin, Will Grathwohl, Daniel Kassler, Jean Disset, Debora S. Marks. Optimal design of stochastic DNA synthesis protocols based on generative sequence models. Artificial Intelligence and Statistics (AISTATS). 2022. paper. code. talk.

Alan N. Amin*, Eli N. Weinstein*, Debora S. Marks (*Equal contribution). A generative nonparametric Bayesian model for whole genomes. Advances in Neural Information Processing Systems (NeurIPS). 2021. paper. code. talk.

Eli N. Weinstein, Debora S. Marks. A structured observation distribution for generative biological sequence prediction and forecasting. Proceedings of the 38th International Conference on Machine Learning (ICML). 2021. paper. Pyro code. Edward2 code. talk.


Eli N. Weinstein. Generative Statistical Methods for Biological Sequences. Harvard University. 2022. pdf


My google scholar page is here.