Expanding functional protein sequence space using generative adversarial networks

By Donatas Repecka, Vykintas Jauniškis, Laurynas Karpus, E. Rembeza, Jan Zrimec, S. Povilonienė, Irmantas Rokaitis, Audrius Laurynėnas, Wissam Abuajwa, Otto Savolainen and 3 more. 2019

105 citations recorded by the source corpus.

De novo protein design for catalysis of any desired chemical reaction is a long standing goal in protein engineering, due to the broad spectrum of technological, scientific and medical applications. Currently, mapping protein sequence to protein function is, however, neither computationionally nor experimentally tangible 1,2. Here we developed ProteinGAN, a specialised variant of the generative adversarial network 3 that is able to ‘learn’ natural protein sequence diversity and enables the generation of functional protein sequences. ProteinGAN learns the evolutionary relationships of protein sequences directly from the complex multidimensional amino acid sequence space and creates new, highly diverse sequence variants with natural-like physical properties. Using malate dehydrogenase as a template enzyme, we show that 24% of the ProteinGAN-generated and experimentally tested sequences are soluble and display wild-type level catalytic activity in the tested conditions in vitro, even in highly mutated (>100 mutations) sequences. ProteinGAN therefore demonstrates the potential of artificial intelligence to rapidly generate highly diverse novel functional proteins within the allowed biological constraints of the sequence space.

Subfields: Molecular Biology, Generative Models, Computational Biology & Bioinformatics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Generative Models

In Computational Biology & Bioinformatics

Citations in the corpus

Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences (2019, background): “It is well known that structural properties of protein sequences are reﬂected in long-range dependencies (Kihara, 2005).”
ProGen2: Exploring the Boundaries of Protein Language Models (2022, background): “Most still train on a ﬁxed protein family to capture coevolutionary signals present within a set of homologous sequences – ranging from direct coupling analysis techniques (Russ et al., 2020) to generative adversarial networks (Repecka et al., 2021).”
A Survey of Deep Learning for Scientific Discovery (2020, background): “GANs have found many scientiﬁc applications, from performing data augmentation in medical image settings [65] to protein generation [193].”
AlphaDesign: A de novo protein design framework based on AlphaFold (2021, background): “Several works train generative models to directly generate protein sequences with a desired function [55–57].”