Toward machine-guided design of proteins

Proteins—molecular machines that underpin all biological life—are of significant therapeutic and industrial value. Directed evolution is a high-throughput experimental approach for improving protein function, but has difficulty escaping local maxima in the fitness landscape. Here, we investigate how supervised learning in a closed loop with DNA synthesis and high-throughput screening can be used to improve protein design. Using the green fluorescent protein (GFP) as an illustrative example, we demonstrate the opportunities and challenges of generating training datasets conducive to selecting strongly generalizing models. With prospectively designed wet lab experiments, we then validate that these models can generalize to unseen regions of the fitness landscape, even when constrained to explore combinations of non-trivial mutations. Taken together, this suggests a hybrid optimization strategy for protein design in which a predictive model is used to explore difficult-to-access but promising regions of the fitness landscape that directed evolution can then exploit at scale.

• Publication year

  2018

• Venue

  bioRxiv

• Publication date

  2018-06-02

• Fields of study

  Biology, Computer Science

• Identifiers
  DOI 10.1101/337154
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• End-to-end differentiable learning of protein structure

  2018cited by this paper
• Machine learning to design integral membrane channelrhodopsins for efficient eukaryotic expression and plasma membrane localization

  2017cited by this paper
• Experimental assay of a fitness landscape on a macroevolutionary scale

  2017cited by this paper
• Self-Normalizing Neural Networks

  2017cited by this paper
• Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach.

  2016cited by this paper
• The coming of age of de novo protein design

  2016cited by this paper
• Protein contact prediction from amino acid co-evolution using convolutional networks for graph-valued images

  2016cited by this paper
• Local fitness landscape of the green fluorescent protein

  2016cited by this paper
• Structural Consequences of Chromophore Formation and Exploration of Conserved Lid Residues amongst Naturally Occurring Fluorescent Proteins.

  2014cited by this paper
• Composability of regulatory sequences controlling transcription and translation in Escherichia coli

  2013cited by this paper
• Fluorescent proteins: shine on, you crazy diamond.

  2013cited by this paper
• Navigating the protein fitness landscape with Gaussian processes

  2012cited by this paper
• Scaling Multidimensional Inference for Structured Gaussian Processes

  2012cited by this paper
• Exploring protein fitness landscapes by directed evolution

  2009cited by this paper
• Engineering proteinase K using machine learning and synthetic genes

  2007cited by this paper
• Improving catalytic function by ProSAR-driven enzyme evolution

  2007cited by this paper
• Engineering and characterization of a superfolder green fluorescent protein

  2006cited by this paper
• Active Learning with Statistical Models

  1996cited by this paper
• Natural Selection and the Concept of a Protein Space

  1970cited by this paper
• Natural Selection and the Concept of a Protein Space

  1970cited by this paper

• Bioengineering hybrid artificial life

  2025influential citation
• Robust Model-Based Optimization for Challenging Fitness Landscapes

  2023cites this paper
• A Transendothelial Leukocyte Transmigration Model Based on Computational Fluid Dynamics and BP Neural Network

  2022cites this paper
• Proximal Exploration for Model-guided Protein Sequence Design

  2022cites this paper
• Randomized gates eliminate bias in sort‐seq assays

  2022cites this paper
• Metabolomics and modelling approaches for systems metabolic engineering

  2022cites this paper
• Rational Design Inspired Application of Natural Language Processing Algorithms to Red Shift mNeptune684

  2021influential citation
• Deep Generative Modeling for Protein Design

  2021cites this paper
• ECNet is an evolutionary context-integrated deep learning framework for protein engineering

  2021cites this paper
• AlphaDesign: A de novo protein design framework based on AlphaFold

  2021cites this paper
• FPredX: Interpretable models for the prediction of spectral maxima, brightness, and oligomeric states of fluorescent proteins

  2021cites this paper
• Improving RNA Secondary Structure Design using Deep Reinforcement Learning

  2021cites this paper
• PyPEF - An Integrated Framework for Data-Driven Protein Engineering

  2021cites this paper
• Heterogeneity of the GFP fitness landscape and data-driven protein design

  2021influential citation
• A machine learning toolkit for genetic engineering attribution to facilitate biosecurity

  2020cites this paper
• Machine learning applications in systems metabolic engineering.

  2020cites this paper
• Evolutionary context-integrated deep sequence modeling for protein engineering

  2020cites this paper
• Low-N protein engineering with data-efficient deep learning

  2020influential citation
• Fast activation maximization for molecular sequence design

  2020cites this paper
• A Hierarchical Approach to Scaling Batch Active Search Over Structured Data

  2020cites this paper
• A Generative Neural Network for Maximizing Fitness and Diversity of Synthetic DNA and Protein Sequences

  2020cites this paper
• Attribution of genetic engineering: A practical and accurate machine-learning toolkit for biosecurity

  2020cites this paper
• AdaLead: A simple and robust adaptive greedy search algorithm for sequence design

  2020cites this paper
• A primer on model-guided exploration of fitness landscapes for biological sequence design

  2020cites this paper
• Neural networks to learn protein sequence–function relationships from deep mutational scanning data

  2020cites this paper
• Directed Evolution of a Selective and Sensitive Serotonin Sensor via Machine Learning.

  2020cites this paper
• Unified rational protein engineering with sequence-based deep representation learning

  2019influential citation
• Deep exploration networks for rapid engineering of functional DNA sequences

  2019cites this paper
• Deep generative models for T cell receptor protein sequences

  2019cites this paper
• Minimum epistasis interpolation for sequence-function relationships

  2019cites this paper
• Leveraging implicit knowledge in neural networks for functional dissection and engineering of proteins

  2019cites this paper
• Unified rational protein engineering with sequence-only deep representation learning

  2019cites this paper
• A Deep Neural Network for Predicting and Engineering Alternative Polyadenylation.

  2019cites this paper
• Generative Models for Graph-Based Protein Design

  2019cites this paper