Exploring amino acid functions in a deep mutational landscape

AI-Predicted Protein Deformation Encodes Energy Landscape Perturbation

AI algorithms have proven to be excellent predictors of protein structure, but whether and how much these algorithms can capture the underlying physics remains an open question. Here, we aim to test this question using the Alphafold2 (AF) algorithm: We use AF to predict the subtle structural deformation induced by single mutations, quantified by strain, and compare with experimental datasets of corresponding perturbations in folding free energy ΔΔG. Unexpectedly, we find that physical strain alone-without any additional data or computation-correlates almost as well with ΔΔG as state-of-the-art energy-based and machine-learning predictors. This indicates that the AF-predicted structures alone encode fine details about the energy landscape. In particular, the structures encode significant information on stability, enough to estimate (de-)stabilizing effects of mutations, thus paving the way for the development of novel, structure-based stability predictors for protein design and evolution.

Deciphering GB1's Single Mutational Landscape: Insights from MuMi Analysis

Mutational changes that affect the binding of the C2 fragment of Streptococcal protein G (GB1) to the Fc domain of human IgG (IgG-Fc) have been extensively studied using deep mutational scanning (DMS), and the binding affinity of all single mutations has been measured experimentally in the literature. To investigate the underlying molecular basis, we perform in silico mutational scanning for all possible single mutations, along with 2 μs-long molecular dynamics (WT-MD) of the wild-type (WT) GB1 in both unbound and IgG-Fc bound forms. We compute the hydrogen bonds between GB1 and IgG-Fc in WT-MD to identify the dominant hydrogen bonds for binding, which we then assess in conformations produced by Mutation and Minimization (MuMi) to explain the fitness landscape of GB1 and IgG-Fc binding. Furthermore, we analyze MuMi and WT-MD to investigate the dynamics of binding, focusing on the relative solvent accessibility of residues and the probability of residues being located at the binding interface. With these analyses, we explain the interactions between GB1 and IgG-Fc and display the structural features of binding. In sum, our findings highlight the potential of MuMi as a reliable and computationally efficient tool for predicting protein fitness landscapes, offering significant advantages over traditional methods. The methodologies and results presented in this study pave the way for improved predictive accuracy in protein stability and interaction studies, which are crucial for advancements in drug design and synthetic biology.

QAFI: a novel method for quantitative estimation of missense variant impact using protein-specific predictors and ensemble learning

Next-generation sequencing (NGS) has revolutionized genetic diagnostics, yet its application in precision medicine remains incomplete, despite significant advances in computational tools for variant annotation. Many variants remain unannotated, and existing tools often fail to accurately predict the range of impacts that variants have on protein function. This limitation restricts their utility in relevant applications such as predicting disease severity and onset age. In response to these challenges, a new generation of computational models is emerging, aimed at producing quantitative predictions of genetic variant impacts. However, the field is still in its early stages, and several issues need to be addressed, including improved performance and better interpretability. This study introduces QAFI, a novel methodology that integrates protein-specific regression models within an ensemble learning framework, utilizing conservation-based and structure-related features derived from AlphaFold models. Our findings indicate that QAFI significantly enhances the accuracy of quantitative predictions across various proteins. The approach has been rigorously validated through its application in the CAGI6 contest, focusing on ARSA protein variants, and further tested on a comprehensive set of clinically labeled variants, demonstrating its generalizability and robust predictive power. The straightforward nature of our models may also contribute to better interpretability of the results.

Machine Learning-Guided Protein Engineering

Recent progress in engineering highly promising biocatalysts has increasingly involved machine learning methods. These methods leverage existing experimental and simulation data to aid in the discovery and annotation of promising enzymes, as well as in suggesting beneficial mutations for improving known targets. The field of machine learning for protein engineering is gathering steam, driven by recent success stories and notable progress in other areas. It already encompasses ambitious tasks such as understanding and predicting protein structure and function, catalytic efficiency, enantioselectivity, protein dynamics, stability, solubility, aggregation, and more. Nonetheless, the field is still evolving, with many challenges to overcome and questions to address. In this Perspective, we provide an overview of ongoing trends in this domain, highlight recent case studies, and examine the current limitations of machine learning-based methods. We emphasize the crucial importance of thorough experimental validation of emerging models before their use for rational protein design. We present our opinions on the fundamental problems and outline the potential directions for future research.

Learning protein fitness landscapes with deep mutational scanning data from multiple sources

One of the key points of machine learning-assisted directed evolution (MLDE) is the accurate learning of the fitness landscape, a conceptual mapping from sequence variants to the desired function. Here, we describe a multi-protein training scheme that leverages the existing deep mutational scanning data from diverse proteins to aid in understanding the fitness landscape of a new protein. Proof-of-concept trials are designed to validate this training scheme in three aspects: random and positional extrapolation for single-variant effects, zero-shot fitness predictions for new proteins, and extrapolation for higher-order variant effects from single-variant effects. Moreover, our study identified previously overlooked strong baselines, and their unexpectedly good performance brings our attention to the pitfalls of MLDE. Overall, these results may improve our understanding of the association between different protein fitness profiles and shed light on developing better machine learning-assisted approaches to the directed evolution of proteins. A record of this paper's transparent peer review process is included in the supplemental information.

Combining deep mutational scanning to heatmap of HLA class II binding of immunogenic sequences to preserve functionality and mitigate predicted immunogenicity

Removal of CD4 T cell epitopes from therapeutic antibody sequences is expected to mitigate their potential immunogenicity, but its application is complicated by the location of their T cell epitopes, which mainly overlap with complementarity-determining regions. We therefore evaluated the flexibility of antibody sequences to reduce the predicted affinity of corresponding peptides for HLA II molecules and to maintain antibody binding to its target in order to guide antibody engineering for mitigation of predicted immunogenicity. Permissive substitutions to reduce affinity of peptides for HLA II molecules were identified by establishing a heatmap of HLA class II binding using T-cell epitope prediction tools, while permissive substitutions preserving binding to the target were identified by means of deep mutational scanning and yeast surface display. Combinatorial libraries were then designed to identify active clones. Applied to adalimumab, an anti-TNFα human antibody, this approach identified 200 mutants with a lower HLA binding score than adalimumab. Three mutants were produced as full-length antibodies and showed a higher affinity for TNFα and neutralization ability than adalimumab. This study also sheds light on the permissiveness of antibody sequences with regard to functionality and predicted T cell epitope content.

EvoRator2: Predicting Site-specific Amino Acid Substitutions Based on Protein Structural Information Using Deep Learning

Multiple sequence alignments (MSAs) are the workhorse of molecular evolution and structural biology research. From MSAs, the amino acids that are tolerated at each site during protein evolution can be inferred. However, little is known regarding the repertoire of tolerated amino acids in proteins when only a few or no sequence homologs are available, such as orphan and de novo designed proteins. Here we present EvoRator2, a deep-learning algorithm trained on over 15,000 protein structures that can predict which amino acids are tolerated at any given site, based exclusively on protein structural information mined from atomic coordinate files. We show that EvoRator2 obtained satisfying results for the prediction of position-weighted scoring matrices (PSSM). We further show that EvoRator2 obtained near state-of-the-art performance on proteins with high quality structures in predicting the effect of mutations in deep mutation scanning (DMS) experiments and that for certain DMS targets, EvoRator2 outperformed state-of-the-art methods. We also show that by combining EvoRator2's predictions with those obtained by a state-of-the-art deep-learning method that accounts for the information in the MSA, the prediction of the effect of mutation in DMS experiments was improved in terms of both accuracy and stability. EvoRator2 is designed to predict which amino-acid substitutions are tolerated in such proteins without many homologous sequences, including orphan or de novo designed proteins. We implemented our approach in the EvoRator web server (https://evorator.tau.ac.il).

Discovering functionally important sites in proteins

Proteins play important roles in biology, biotechnology and pharmacology, and missense variants are a common cause of disease. Discovering functionally important sites in proteins is a central but difficult problem because of the lack of large, systematic data sets. Sequence conservation can highlight residues that are functionally important but is often convoluted with a signal for preserving structural stability. We here present a machine learning method to predict functional sites by combining statistical models for protein sequences with biophysical models of stability. We train the model using multiplexed experimental data on variant effects and validate it broadly. We show how the model can be used to discover active sites, as well as regulatory and binding sites. We illustrate the utility of the model by prospective prediction and subsequent experimental validation on the functional consequences of missense variants in HPRT1 which may cause Lesch-Nyhan syndrome, and pinpoint the molecular mechanisms by which they cause disease.

High-throughput deep learning variant effect prediction with Sequence UNET

Understanding coding mutations is important for many applications in biology and medicine but the vast mutation space makes comprehensive experimental characterisation impossible. Current predictors are often computationally intensive and difficult to scale, including recent deep learning models. We introduce Sequence UNET, a highly scalable deep learning architecture that classifies and predicts variant frequency from sequence alone using multi-scale representations from a fully convolutional compression/expansion architecture. It achieves comparable pathogenicity prediction to recent methods. We demonstrate scalability by analysing 8.3B variants in 904,134 proteins detected through large-scale proteomics. Sequence UNET runs on modest hardware with a simple Python package.

A complete allosteric map of a GTPase switch in its native cellular network

Allosteric regulation is central to protein function in cellular networks. A fundamental open question is whether cellular regulation of allosteric proteins occurs only at a few defined positions or at many sites distributed throughout the structure. Here, we probe the regulation of GTPases-protein switches that control signaling through regulated conformational cycling-at residue-level resolution by deep mutagenesis in the native biological network. For the GTPase Gsp1/Ran, we find that 28% of the 4,315 assayed mutations show pronounced gain-of-function responses. Twenty of the sixty positions enriched for gain-of-function mutations are outside the canonical GTPase active site switch regions. Kinetic analysis shows that these distal sites are allosterically coupled to the active site. We conclude that the GTPase switch mechanism is broadly sensitive to cellular allosteric regulation. Our systematic discovery of new regulatory sites provides a functional map to interrogate and target GTPases controlling many essential biological processes.

Deep mutational scanning of essential bacterial proteins can guide antibiotic development

Deep mutational scanning is a powerful approach to investigate a wide variety of research questions including protein function and stability. Here, we perform deep mutational scanning on three essential E. coli proteins (FabZ, LpxC and MurA) involved in cell envelope synthesis using high-throughput CRISPR genome editing, and study the effect of the mutations in their original genomic context. We use more than 17,000 variants of the proteins to interrogate protein function and the importance of individual amino acids in supporting viability. Additionally, we exploit these libraries to study resistance development against antimicrobial compounds that target the selected proteins. Among the three proteins studied, MurA seems to be the superior antimicrobial target due to its low mutational flexibility, which decreases the chance of acquiring resistance-conferring mutations that simultaneously preserve MurA function. Additionally, we rank anti-LpxC lead compounds for further development, guided by the number of resistance-conferring mutations against each compound. Our results show that deep mutational scanning studies can be used to guide drug development, which we hope will contribute towards the development of novel antimicrobial therapies.

Deciphering cross-species reactivity of LAMP-1 antibodies using deep mutational epitope mapping and AlphaFold

Delineating the precise regions on an antigen that are targeted by antibodies has become a key step for the development of antibody therapeutics. X-ray crystallography and cryogenic electron microscopy are considered the gold standard for providing precise information about these binding sites at atomic resolution. However, they are labor-intensive and a successful outcome is not guaranteed. We used deep mutational scanning (DMS) of the human LAMP-1 antigen displayed on yeast surface and leveraged next-generation sequencing to observe the effect of individual mutants on the binding of two LAMP-1 antibodies and to determine their functional epitopes on LAMP-1. Fine-tuned epitope mapping by DMS approaches is augmented by knowledge of experimental antigen structure. As human LAMP-1 structure has not yet been solved, we used the AlphaFold predicted structure of the full-length protein to combine with DMS data and ultimately finely map antibody epitopes. The accuracy of this method was confirmed by comparing the results to the co-crystal structure of one of the two antibodies with a LAMP-1 luminal domain. Finally, we used AlphaFold models of non-human LAMP-1 to understand the lack of mAb cross-reactivity. While both epitopes in the murine form exhibit multiple mutations in comparison to human LAMP-1, only one and two mutations in the Macaca form suffice to hinder the recognition by mAb B and A, respectively. Altogether, this study promotes a new application of AlphaFold to speed up precision mapping of antibody-antigen interactions and consequently accelerate antibody engineering for optimization.

A structural biology community assessment of AlphaFold2 applications

Most proteins fold into 3D structures that determine how they function and orchestrate the biological processes of the cell. Recent developments in computational methods for protein structure predictions have reached the accuracy of experimentally determined models. Although this has been independently verified, the implementation of these methods across structural-biology applications remains to be tested. Here, we evaluate the use of AlphaFold2 (AF2) predictions in the study of characteristic structural elements; the impact of missense variants; function and ligand binding site predictions; modeling of interactions; and modeling of experimental structural data. For 11 proteomes, an average of 25% additional residues can be confidently modeled when compared with homology modeling, identifying structural features rarely seen in the Protein Data Bank. AF2-based predictions of protein disorder and complexes surpass dedicated tools, and AF2 models can be used across diverse applications equally well compared with experimentally determined structures, when the confidence metrics are critically considered. In summary, we find that these advances are likely to have a transformative impact in structural biology and broader life-science research.

Intragenic compensation through the lens of deep mutational scanning

A significant fraction of mutations in proteins are deleterious and result in adverse consequences for protein function, stability, or interaction with other molecules. Intragenic compensation is a specific case of positive epistasis when a neutral missense mutation cancels effect of a deleterious mutation in the same protein. Permissive compensatory mutations facilitate protein evolution, since without them all sequences would be extremely conserved. Understanding compensatory mechanisms is an important scientific challenge at the intersection of protein biophysics and evolution. In human genetics, intragenic compensatory interactions are important since they may result in variable penetrance of pathogenic mutations or fixation of pathogenic human alleles in orthologous proteins from related species. The latter phenomenon complicates computational and clinical inference of an allele's pathogenicity. Deep mutational scanning is a relatively new technique that enables experimental studies of functional effects of thousands of mutations in proteins. We review the important aspects of the field and discuss existing limitations of current datasets. We reviewed ten published DMS datasets with quantified functional effects of single and double mutations and described rates and patterns of intragenic compensation in eight of them.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12551-022-01005-w.

Redesigning Enzymes for Biocatalysis: Exploiting Structural Understanding for Improved Selectivity

The discovery of new enzymes, alongside the push to make chemical processes more sustainable, has resulted in increased industrial interest in the use of biocatalytic processes to produce high-value and chiral precursor chemicals. Huge strides in protein engineering methodology and in silico tools have facilitated significant progress in the discovery and production of enzymes for biocatalytic processes. However, there are significant gaps in our knowledge of the relationship between enzyme structure and function. This has demonstrated the need for improved computational methods to model mechanisms and understand structure dynamics. Here, we explore efforts to rationally modify enzymes toward changing aspects of their catalyzed chemistry. We highlight examples of enzymes where links between enzyme function and structure have been made, thus enabling rational changes to the enzyme structure to give predictable chemical outcomes. We look at future directions the field could take and the technologies that will enable it.

Recent Advances in Machine Learning Variant Effect Prediction Tools for Protein Engineering

Proteins are Nature's molecular machinery and comprise diverse roles while consisting of chemically similar building blocks. In recent years, protein engineering and design have become important research areas, with many applications in the pharmaceutical, energy, and biocatalysis fields, among others-where the aim is to ultimately create a protein given desired structural and functional properties. It is often critical to model the relationship between a protein's sequence, folded structure, and biological function to assist in such protein engineering pursuits. However, significant challenges remain in concretely mapping an amino acid sequence to specific protein properties and biological activities. Mutations may enhance or diminish molecular protein function, and the epistatic interactions between mutations result in an inherently complex mapping between genetic modifications and protein function. Therefore, estimating the quantitative effects of mutations on protein function(s) remains a grand challenge of biology, bioinformatics, and many related fields and would rapidly accelerate protein engineering tasks when successful. Such estimation is often known as variant effect prediction (VEP). However, progress has been demonstrated in recent years with the development of machine learning (ML) methods in modeling the relationship between mutations and protein function. In this Review, recent advances in variant effect prediction (VEP) are discussed as tools for protein engineering, focusing on techniques incorporating gains from the broader ML community and challenges in estimating biomolecular functional differences. Primary developments highlighted include convolutional neural networks, graph neural networks, and natural language embeddings for protein sequences.

Variant Library Annotation Tool (VaLiAnT): an oligonucleotide library design and annotation tool for saturation genome editing and other deep mutational scanning experiments

MOTIVATION

CRISPR/Cas9-based technology allows for the functional analysis of genetic variants at single nucleotide resolution whilst maintaining genomic context. This approach, known as saturation genome editing (SGE), a form of deep mutational scanning, systematically alters each position in a target region to explore its function. SGE experiments require the design and synthesis of oligonucleotide variant libraries which are introduced into the genome. This technology is applicable to diverse fields such as disease variant identification, drug development, structure-function studies, synthetic biology, evolutionary genetics and host-pathogen interactions. Here, we present the Variant Library Annotation Tool (VaLiAnT) which can be used to generate variant libraries from user-defined genomic coordinates and standard input files. The software can accommodate user-specified species, reference sequences and transcript annotations.

RESULTS

Coordinates for a genomic range are provided by the user to retrieve a corresponding oligonucleotide reference sequence. A user-specified range within this sequence is then subject to systematic, nucleotide and/or amino acid saturating mutator functions. VaLiAnT provides a novel way to retrieve, mutate and annotate genomic sequences for oligonucleotide library generation. Specific features for SGE library generation can be employed. In addition, VaLiAnT is configurable, allowing for cDNA and prime editing saturation library generation, with other diverse applications possible.

AVAILABILITY AND IMPLEMENTATION

VaLiAnT is a command line tool written in Python. Source code, testing data, example input and output files and executables are available (https://github.com/cancerit/VaLiAnT) in addition to a detailed user manual (https://github.com/cancerit/VaLiAnT/wiki). VaLiAnT is licensed under AGPLv3.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Predicting and interpreting large-scale mutagenesis data using analyses of protein stability and conservation

Understanding and predicting the functional consequences of single amino acid changes is central in many areas of protein science. Here, we collect and analyze experimental measurements of effects of >150,000 variants in 29 proteins. We use biophysical calculations to predict changes in stability for each variant and assess them in light of sequence conservation. We find that the sequence analyses give more accurate prediction of variant effects than predictions of stability and that about half of the variants that show loss of function do so due to stability effects. We construct a machine learning model to predict variant effects from protein structure and sequence alignments and show how the two sources of information support one another and enable mechanistic interpretations. Together, our results show how one can leverage large-scale experimental assessments of variant effects to gain deeper and general insights into the mechanisms that cause loss of function.

Facilitating Machine Learning-Guided Protein Engineering with Smart Library Design and Massively Parallel Assays

Protein design plays an important role in recent medical advances from antibody therapy to vaccine design. Typically, exhaustive mutational screens or directed evolution experiments are used for the identification of the best design or for improvements to the wild-type variant. Even with a high-throughput screening on pooled libraries and Next-Generation Sequencing to boost the scale of read-outs, surveying all the variants with combinatorial mutations for their empirical fitness scores is still of magnitudes beyond the capacity of existing experimental settings. To tackle this challenge, in-silico approaches using machine learning to predict the fitness of novel variants based on a subset of empirical measurements are now employed. These machine learning models turn out to be useful in many cases, with the premise that the experimentally determined fitness scores and the amino-acid descriptors of the models are informative. The machine learning models can guide the search for the highest fitness variants, resolve complex epistatic relationships, and highlight bio-physical rules for protein folding. Using machine learning-guided approaches, researchers can build more focused libraries, thus relieving themselves from labor-intensive screens and fast-tracking the optimization process. Here, we describe the current advances in massive-scale variant screens, and how machine learning and mutagenesis strategies can be integrated to accelerate protein engineering. More specifically, we examine strategies to make screens more economical, informative, and effective in discovery of useful variants.

Understanding the Origins of Loss of Protein Function by Analyzing the Effects of Thousands of Variants on Activity and Abundance

Understanding and predicting how amino acid substitutions affect proteins are keys to our basic understanding of protein function and evolution. Amino acid changes may affect protein function in a number of ways including direct perturbations of activity or indirect effects on protein folding and stability. We have analyzed 6,749 experimentally determined variant effects from multiplexed assays on abundance and activity in two proteins (NUDT15 and PTEN) to quantify these effects and find that a third of the variants cause loss of function, and about half of loss-of-function variants also have low cellular abundance. We analyze the structural and mechanistic origins of loss of function and use the experimental data to find residues important for enzymatic activity. We performed computational analyses of protein stability and evolutionary conservation and show how we may predict positions where variants cause loss of activity or abundance. In this way, our results link thermodynamic stability and evolutionary conservation to experimental studies of different properties of protein fitness landscapes.