Chloe Hsu

I build tools for the design and interpretation of proteins. This includes machine learning methods for structure-based protein design and for sequence ranking in protein engineering.

This was my academic website. I completed my PhD at the University of California, Berkeley, advised by Jennifer Listgarten and Moritz Hardt.

Research

Generative models for protein structures and sequences

(Published as a primer) How can generative models be useful for protein engineering?

Learning inverse folding from millions of predicted structures

Inverse folding aims to design sequences to fold into desired structures. With 12M new predicted structures as additional training data, ESM-IF1 is more accurate at structure-based sequence design, while also generalizing to more sophisticated design tasks.

Learning protein fitness models from evolutionary and assay-labeled data

A simple yet highly effective hybrid approach to protein fitness prediction. Also a comparative analysis to highlight the importance of systematic evaluations and sufficient baselines.

Nanopore callers for epigenetics from limited supervised data

Calling epigenetic modifications on nanopore sequencing platforms when the training data is incomplete.

Design and source code from Jon Barron's website and Leonid Keselman's Jekyll fork.