Biological Design and Engineering By and for machine Learning | AIChE

Other Sites & Tools

Technical Groups

Biological Design and Engineering By and for machine Learning

Authors 

Faulon, J. L. - Presenter, Genoscope
Faure, L., INRAe, University of Paris Saclay
Duigou, T., INRA
Kushwaha, M., Institut national de la recherche agronomique (INRA)
We have seen the past few years a growing interest in using machine learning in chemistry and biology and synthetic biology makes no exception to this trend. Among, the main learning methods found in the synthetic biology literature are supervised learning, reinforcement and active learning, and in vivo/in vitro learning.

Supervised machine learning is being exploited to predict sequence activities, engineer sequences, associate biological signals with phenotypes, and optimize culture conditions. An example of promiscuous enzyme activity prediction with Gaussian processes and Deep Learning will be shown in the context of underground metabolism and metabolome completion [1].

Reinforcement and active learning are using training sets acquired through an iterative process. These methods are particularly amendable to the Design-Build-Test-Learn synthetic biology cycle. Reinforcement learning will be exemplified for bio-retrosynthesis [2] and active learning to maximize the productivity of cell-free systems [3].

Engineering information processing devices in living systems is a long-standing venture of synthetic biology. Yet, the problem of engineering devices that perform basic operations found in machine learning remains largely unexplored. As a first step toward engineering biological learning devices, a metabolic perceptron will be presented. The performances of the perceptron will be exemplified with biological sample classification based on metabolic composition [4].

References:

1. Mellor J, et al. Semi-supervised Gaussian Process for automated enzyme search. ACS Synthetic Biology, 2016, 5(6): 518-528.

2. Koch M, et al. Reinforcement Learning for Bioretrosynthesis. ACS Synthetic Biology, 2020, 9(1): 157-168.

3. Borkowski O, et al. Large scale active-learning-guided exploration to maximize cell-free production, Nature Communications, 2020, 11: 1872

4. Pandi A, et al. Metabolic Perceptrons for Neural Computing in Biological Systems. Nature Communications, 2019, 10: 3880.

More Conference Links