MIA: Eric Kelsic, Machine-guided capsid engineering for gene therapy; Sam Sinai, Sequence design

October 7, 2020
Models, Inference and Algorithms
Broad Institute

Machine-guided capsid engineering for gene therapy

Eric Kelsic
Dyno Therapeutics

Machine-guided capsid engineering represents a new approach to overcome the current challenges of in vivo gene delivery. Such an approach combines three advanced technologies: i. next-gen library synthesis, ii. next-gen sequencing, and iii. machine learning. With this workflow the search for improved capsids can be dramatically accelerated. This talk will review the technological advances that are pushing the field of AAV capsid engineering toward machine-guided methods, describe and explore the promise of this new approach, and discuss anticipated challenges. In the near future, machine-guided methods will revolutionize our ability to design robustly efficient, safe and targeted vectors for the treatment of genetic conditions.

Primer: Biological sequence design through machine-guided exploration

Sam Sinai
Dyno Therapeutics

Efficient design of biological sequences will have a great impact across many industry and healthcare domains. However, discovering improved sequences requires solving a difficult optimization problem. Traditionally, this challenge was approached by biologists through a model-free method known as “directed evolution”, the iterative process of random mutation and selection. As the ability to build models that capture the sequence-to-function map improves, such models can be used as oracles to screen sequences before committing to experiments. In recent years, interest in better algorithms that effectively use such oracles to outperform model-free approaches has intensified. These span from standard Bayesian Optimization approaches, to regularized generative models and adaptations of reinforcement learning. This primer will compare such algorithms based on a comprehensive set of criteria that are important from both machine learning and biological perspectives.

Copyright Broad Institute, 2020. All rights reserved.