Our Mission

Recent advances in high-throughput experimental methods and machine learning approaches have fueled interest in ML-driven protein design. These advances may enable more rapid development of designed proteins with applications ranging from biopharmaceuticals, catalysis, material design and basic science research. However, this excitement has exposed important research questions across the foundation of this emerging engineering discipline.

For example:

• What experimental approaches can feed the data-driven design cycle?

• Which machine learning models and parameterizations of proteins hold the right inductive biases?

• What are the limits of the growing structural and evolutionary data in the PDB and UniProt?

• How do we use our trained models to guide data collection?

We think these questions will be best addressed by a collaborative, interdisciplinary community. Thus, the ML4Protein Engineering community runs a bi-weekly seminar series to address these advances and other outstanding problems, such as high-throughput screening, model-based optimization, and representation learning.

To access announcements, please follow us on Twitteror Bluesky! You can also visit our YouTube to see recordings of past talks! Also, be sure to join ourSlack Community, where we discuss even more opportunities beyond the seminar series!

Check out our new seminar series, Early Career Scientists, below!

Meg Taylor
Duke Biomedical Engineering PhD Student

Sonia Yuan
CalTech Biochemistry & Molecular Biophysics PhD student

Leo Zang
Harvard Medical School SSQBio PhD Student

Bo Qiang
University of Washington Computer Science PhD Student