On Simplified Global Nonlinear Function for Fitness Landscape: A Case Study of Inverse Protein Folding

The construction of fitness landscape has broad implication in understanding molecular evolution, cellular epigenetic state, and protein structures. We studied the problem of constructing fitness landscape of inverse protein folding or protein design, with the aim to generate amino acid sequences that would fold into an a priori determined structural fold which would enable engineering novel or enhanced biochemistry. For this task, an effective fitness function should allow identification of correct sequences that would fold into the desired structure. In this study, we showed that nonlinear fitness function for protein design can be constructed using a rectangular kernel with a basis set of proteins and decoys chosen a priori. The full landscape for a large number of protein folds can be captured using only 480 native proteins and 3,200 non-protein decoys via a finite Newton method. A blind test of a simplified version of fitness function for sequence design was carried out to discriminate simultaneously 428 native sequences not homologous to any training proteins from 11 million challenging protein-like decoys. This simplified function correctly classified 408 native sequences (20 misclassifications, 95% correct rate), which outperforms several other statistical linear scoring function and optimized linear function. Our results further suggested that for the task of global sequence design of 428 selected proteins, the search space of protein shape and sequence can be effectively parametrized with just about 3,680 carefully chosen basis set of proteins and decoys, and we showed in addition that the overall landscape is not overly sensitive to the specific choice of this set. Our results can be generalized to construct other types of fitness landscape.     

Generating functional protein variants with variational autoencoders

The vast expansion of protein sequence databases provides an opportunity for new protein design approaches which seek to learn the sequence-function relationship directly from natural sequence variation. Deep generative models trained on protein sequence data have been shown to learn biologically meaningful representations helpful for a variety of downstream tasks, but their potential for direct use in the design of novel proteins remains largely unexplored. Here we show that variational autoencoders trained on a dataset of almost 70000 luciferase-like oxidoreductases can be used to generate novel, functional variants of the luxA bacterial luciferase. We propose separate VAE models to work with aligned sequence input (MSA VAE) and raw sequence input (AR-VAE), and offer evidence that while both are able to reproduce patterns of amino acid usage characteristic of the family, the MSA VAE is better able to capture long-distance dependencies reflecting the influence of 3D structure. To confirm the practical utility of the models, we used them to generate variants of luxA whose luminescence activity was validated experimentally. We further showed that conditional variants of both models could be used to increase the solubility of luxA without disrupting function. Altogether 6/12 of the variants generated using the unconditional AR-VAE and 9/11 generated using the unconditional MSA VAE retained measurable luminescence, together with all 23 of the less distant variants generated by conditional versions of the models; the most distant functional variant contained 35 differences relative to the nearest training set sequence. These results demonstrate the feasibility of using deep generative models to explore the space of possible protein sequences and generate useful variants, providing a method complementary to rational design and directed evolution approaches.     

Predicting mostly disordered proteins by using structure-unknown protein data

Background  

Predicting intrinsically disordered proteins is important in structural biology because they are thought to carry out various cellular functions even though they have no stable three-dimensional structure. We know the structures of far more ordered proteins than disordered proteins. The structural distribution of proteins in nature can therefore be inferred to differ from that of proteins whose structures have been determined experimentally. We know many more protein sequences than we do protein structures, and many of the known sequences can be expected to be those of disordered proteins. Thus it would be efficient to use the information of structure-unknown proteins in order to avoid training data sparseness. We propose a novel method for predicting which proteins are mostly disordered by using spectral graph transducer and training with a huge amount of structure-unknown sequences as well as structure-known sequences.     

An Evolution-Based Approach to De Novo Protein Design and Case Study on Mycobacterium tuberculosis

Computational protein design is a reverse procedure of protein folding and structure prediction, where constructing structures from evolutionarily related proteins has been demonstrated to be the most reliable method for protein 3-dimensional structure prediction. Following this spirit, we developed a novel method to design new protein sequences based on evolutionarily related protein families. For a given target structure, a set of proteins having similar fold are identified from the PDB library by structural alignments. A structural profile is then constructed from the protein templates and used to guide the conformational search of amino acid sequence space, where physicochemical packing is accommodated by single-sequence based solvation, torsion angle, and secondary structure predictions. The method was tested on a computational folding experiment based on a large set of 87 protein structures covering different fold classes, which showed that the evolution-based design significantly enhances the foldability and biological functionality of the designed sequences compared to the traditional physics-based force field methods. Without using homologous proteins, the designed sequences can be folded with an average root-mean-square-deviation of 2.1 Å to the target. As a case study, the method is extended to redesign all 243 structurally resolved proteins in the pathogenic bacteria Mycobacterium tuberculosis, which is the second leading cause of death from infectious disease. On a smaller scale, five sequences were randomly selected from the design pool and subjected to experimental validation. The results showed that all the designed proteins are soluble with distinct secondary structure and three have well ordered tertiary structure, as demonstrated by circular dichroism and NMR spectroscopy. Together, these results demonstrate a new avenue in computational protein design that uses knowledge of evolutionary conservation from protein structural families to engineer new protein molecules of improved fold stability and biological functionality.     

Design of multispecific protein sequences using probabilistic graphical modeling

In nature, proteins partake in numerous protein– protein interactions that mediate their functions. Moreover, proteins have been shown to be physically stable in multiple structures, induced by cellular conditions, small ligands, or covalent modifications. Understanding how protein sequences achieve this structural promiscuity at the atomic level is a fundamental step in the drug design pipeline and a critical question in protein physics. One way to investigate this subject is to computationally predict protein sequences that are compatible with multiple states, i.e., multiple target structures or binding to distinct partners. The goal of engineering such proteins has been termed multispecific protein design. We develop a novel computational framework to efficiently and accurately perform multispecific protein design. This framework utilizes recent advances in probabilistic graphical modeling to predict sequences with low energies in multiple target states. Furthermore, it is also geared to specifically yield positional amino acid probability profiles compatible with these target states. Such profiles can be used as input to randomly bias high‐throughput experimental sequence screening techniques, such as phage display, thus providing an alternative avenue for elucidating the multispecificity of natural proteins and the synthesis of novel proteins with specific functionalities. We prove the utility of such multispecific design techniques in better recovering amino acid sequence diversities similar to those resulting from millions of years of evolution. We then compare the approaches of prediction of low energy ensembles and of amino acid profiles and demonstrate their complementarity in providing more robust predictions for protein design. Proteins 2010. © 2009 Wiley‐Liss, Inc.     

Towards Prediction of Metabolic Products of Polyketide Synthases: An In Silico Analysis

Sequence data arising from an increasing number of partial and complete genome projects is revealing the presence of the polyketide synthase (PKS) family of genes not only in microbes and fungi but also in plants and other eukaryotes. PKSs are huge multifunctional megasynthases that use a variety of biosynthetic paradigms to generate enormously diverse arrays of polyketide products that posses several pharmaceutically important properties. The remarkable conservation of these gene clusters across organisms offers abundant scope for obtaining novel insights into PKS biosynthetic code by computational analysis. We have carried out a comprehensive in silico analysis of modular and iterative gene clusters to test whether chemical structures of the secondary metabolites can be predicted from PKS protein sequences. Here, we report the success of our method and demonstrate the feasibility of deciphering the putative metabolic products of uncharacterized PKS clusters found in newly sequenced genomes. Profile Hidden Markov Model analysis has revealed distinct sequence features that can distinguish modular PKS proteins from their iterative counterparts. For iterative PKS proteins, structural models of iterative ketosynthase (KS) domains have revealed novel correlations between the size of the polyketide products and volume of the active site pocket. Furthermore, we have identified key residues in the substrate binding pocket that control the number of chain extensions in iterative PKSs. For modular PKS proteins, we describe for the first time an automated method based on crucial intermolecular contacts that can distinguish the correct biosynthetic order of substrate channeling from a large number of non-cognate combinatorial possibilities. Taken together, our in silico analysis provides valuable clues for formulating rules for predicting polyketide products of iterative as well as modular PKS clusters. These results have promising potential for discovery of novel natural products by genome mining and rational design of novel natural products.     

Developing optimal non-linear scoring function for protein design

Motivation. Protein design aims to identify sequences compatible with a given protein fold but incompatible to any alternative folds. To select the correct sequences and to guide the search process, a design scoring function is critically important. Such a scoring function should be able to characterize the global fitness landscape of many proteins simultaneously. RESULTS: To find optimal design scoring functions, we introduce two geometric views and propose a formulation using a mixture of non-linear Gaussian kernel functions. We aim to solve a simplified protein sequence design problem. Our goal is to distinguish each native sequence for a major portion of representative protein structures from a large number of alternative decoy sequences, each a fragment from proteins of different folds. Our scoring function discriminates perfectly a set of 440 native proteins from 14 million sequence decoys. We show that no linear scoring function can succeed in this task. In a blind test of unrelated proteins, our scoring function misclassfies only 13 native proteins out of 194. This compares favorably with about three-four times more misclassifications when optimal linear functions reported in the literature are used. We also discuss how to develop protein folding scoring function.     

Thoroughly sampling sequence space: large-scale protein design of structural ensembles

Modeling the inherent flexibility of the protein backbone as part of computational protein design is necessary to capture the behavior of real proteins and is a prerequisite for the accurate exploration of protein sequence space. We present the results of a broad exploration of sequence space, with backbone flexibility, through a novel approach: large-scale protein design to structural ensembles. A distributed computing architecture has allowed us to generate hundreds of thousands of diverse sequences for a set of 253 naturally occurring proteins, allowing exciting insights into the nature of protein sequence space. Designing to a structural ensemble produces a much greater diversity of sequences than previous studies have reported, and homology searches using profiles derived from the designed sequences against the Protein Data Bank show that the relevance and quality of the sequences is not diminished. The designed sequences have greater overall diversity than corresponding natural sequence alignments, and no direct correlations are seen between the diversity of natural sequence alignments and the diversity of the corresponding designed sequences. For structures in the same fold, the sequence entropies of the designed sequences cluster together tightly. This tight clustering of sequence entropies within a fold and the separation of sequence entropy distributions for different folds suggest that the diversity of designed sequences is primarily determined by a structure's overall fold, and that the designability principle postulated from studies of simple models holds in real proteins. This has important implications for experimental protein design and engineering, as well as providing insight into protein evolution.     

Deep learning for reconstructing protein structures from cryo-EM density maps: Recent advances and future directions

Cryo-Electron Microscopy (cryo-EM) has emerged as a key technology to determine the structure of proteins, particularly large protein complexes and assemblies in recent years. A key challenge in cryo-EM data analysis is to automatically reconstruct accurate protein structures from cryo-EM density maps. In this review, we briefly overview various deep learning methods for building protein structures from cryo-EM density maps, analyze their impact, and discuss the challenges of preparing high-quality data sets for training deep learning models. Looking into the future, more advanced deep learning models of effectively integrating cryo-EM data with other sources of complementary data such as protein sequences and AlphaFold-predicted structures need to be developed to further advance the field.     

基于深度学习的蛋白质建模与设计

随着蛋白质序列及结构数据的大量累积,在获得了大量描述性信息之后如何有效利用海量数据,从已有数据中高效提取信息并且应用到下游任务当中就成为了研究者亟待解决的问题。蛋白质的设计可使新蛋白的研发不再受限于实验条件,这对药物靶点预测、新药研发和材料设计等领域具有重要意义。深度学习作为一种高效的数据特征提取方法,可以通过它对蛋白质数据进行建模,进而加入先验信息对蛋白质进行设计。故此基于深度学习的蛋白质设计就成为一个具有广阔前景的研究领域。文中主要阐述基于深度学习的蛋白质序列与结构数据的建模和设计方法。详述该方法的策略、原理、适用范围、应用实例。讨论了深度学习方法在本领域的应用前景及局限性,以期为相关研究提供参考。     

Multiple Profile Models Extract Features from Protein Sequence Data and Resolve Functional Diversity of Very Different Protein Families

Functional classification of proteins from sequences alone has become a critical bottleneck in understanding the myriad of protein sequences that accumulate in our databases. The great diversity of homologous sequences hides, in many cases, a variety of functional activities that cannot be anticipated. Their identification appears critical for a fundamental understanding of the evolution of living organisms and for biotechnological applications. ProfileView is a sequence-based computational method, designed to functionally classify sets of homologous sequences. It relies on two main ideas: the use of multiple profile models whose construction explores evolutionary information in available databases, and a novel definition of a representation space in which to analyze sequences with multiple profile models combined together. ProfileView classifies protein families by enriching known functional groups with new sequences and discovering new groups and subgroups. We validate ProfileView on seven classes of widespread proteins involved in the interaction with nucleic acids, amino acids and small molecules, and in a large variety of functions and enzymatic reactions. ProfileView agrees with the large set of functional data collected for these proteins from the literature regarding the organization into functional subgroups and residues that characterize the functions. In addition, ProfileView resolves undefined functional classifications and extracts the molecular determinants underlying protein functional diversity, showing its potential to select sequences towards accurate experimental design and discovery of novel biological functions. On protein families with complex domain architecture, ProfileView functional classification reconciles domain combinations, unlike phylogenetic reconstruction. ProfileView proves to outperform the functional classification approach PANTHER, the two k-mer-based methods CUPP and eCAMI and a neural network approach based on Restricted Boltzmann Machines. It overcomes time complexity limitations of the latter.     

Computational protein design promises to revolutionize protein engineering

Natural evolution has produced an astounding array of proteins that perform the physical and chemical functions required for life on Earth. Although proteins can be reengineered to provide altered or novel functions, the utility of this approach is limited by the difficulty of identifying protein sequences that display the desired properties. Recently, advances in the field of computational protein design (CPD) have shown that molecular simulation can help to predict sequences with new and improved functions. In the past few years, CPD has been used to design protein variants with optimized specificity of binding to DNA, small molecules, peptides, and other proteins. Initial successes in enzyme design highlight CPD's unique ability to design function de novo. The use of CPD for the engineering of potential therapeutic agents has demonstrated its strength in real-life applications.     

The use of soluble protein structures in modeling helical proteins in a layered membrane

Major advances have been made in the prediction of soluble protein structures, led by the knowledge-based modeling methods that extract useful structural trends from known protein structures and incorporate them into scoring functions. The same cannot be reported for the class of transmembrane proteins, primarily due to the lack of high-resolution structural data for transmembrane proteins, which render many of the knowledge-based method unreliable or invalid. We have developed a method that harnesses the vast structural knowledge available in soluble protein data for use in the modeling of transmembrane proteins. At the core of the method, a set of transmembrane protein decoy sets that allow us to filter and train features recognized from soluble proteins for transmembrane protein modeling into a set of scoring functions. We have demonstrated that structures of soluble proteins can provide significant insight into transmembrane protein structures. A complementary novel two-stage modeling/selection process that mimics the two-stage helical membrane protein folding was developed. Combined with the scoring function, the method was successfully applied to model 5 transmembrane proteins. The root mean square deviations of the predicted models ranged from 5.0 to 8.8?Å to the native structures.     

The opportunities and challenges posed by the new generation of deep learning-based protein structure predictors

The function of proteins can often be inferred from their three-dimensional structures. Experimental structural biologists spent decades studying these structures, but the accelerated pace of protein sequencing continuously increases the gaps between sequences and structures. The early 2020s saw the advent of a new generation of deep learning-based protein structure prediction tools that offer the potential to predict structures based on any number of protein sequences.In this review, we give an overview of the impact of this new generation of structure prediction tools, with examples of the impacted field in the life sciences. We discuss the novel opportunities and new scientific and technical challenges these tools present to the broader scientific community. Finally, we highlight some potential directions for the future of computational protein structure prediction.     

A particular set of small non-coding RNAs is bound to the distinctive Argonaute protein of Trypanosoma cruzi: Insights from RNA-interference deficient organisms

The study of small RNAs and Argonaute proteins in eukaryotes that are deficient in functional RNA interference could provide insights into novel functions of small RNAs. In this study we describe small non-coding RNAs bound to a distinctive Argonaute protein of Trypanosoma cruzi, TcPIWI-tryp. Co-immunoprecipitation of TcPIWI-tryp followed by deep sequencing of isolated RNA identified abundant small RNAs derived from rRNAs and tRNAs. The small RNA repertoire differed from that of the canonical Argonaute in organisms with functional RNA interference, which could indicate novel biological functions for TcPIWI-tryp in T. cruzi and other members of the trypanosomatid clade.     

Engineering by homologous recombination: exploring sequence and function within a conserved fold

In nature similar protein folds accommodate distant sequences and support diverse functions. This observation coupled with the recognition that proteins can tolerate many homologous substitutions inspires protein engineers to use recombination to search for new functions within sequences encoding structurally related molecules. These searches have led to proteins with novel activities, diversified specificities and greater stabilities. Computational methods that exploit structural and evolutionary information are being used to design highly mutated yet still natively folded chimeric proteins and protein libraries.     

Reconstruction and function of ancestral center-of-tree human immunodeficiency virus type 1 proteins

The extensive diversity of human immunodeficiency virus type 1 (HIV-1) and its capacity to mutate and escape host immune responses are major challenges for AIDS vaccine development. Ancestral sequences, which minimize the genetic distance to circulating strains, provide an opportunity to design immunogens with the potential to elicit broad recognition of HIV epitopes. We developed a phylogenetics-informed algorithm to reconstruct ancestral HIV sequences, called Center of Tree (COT). COT sequences have potentially significant benefits over isolate-based strategies, as they minimize the evolutionary distances to circulating strains. COT sequences are designed to surmount the potential pitfalls stemming from sampling bias with the consensus method and outlier bias with the most-recent-common-ancestor approach. We computationally derived COT sequences from circulating HIV-1 subtype B sequences for the genes encoding the major viral structural protein (Gag) and two regulatory proteins, Tat and Nef. COT genes were synthesized de novo and expressed in mammalian cells, and the proteins were characterized. COT Gag was shown to generate virus-like particles, while COT Tat transactivated gene expression from the HIV-1 long terminal repeat and COT Nef mediated downregulation of cell surface major histocompatibility complex class I. Thus, retrodicted ancestral COT proteins can retain the biological functions of extant HIV-1 proteins. Additionally, COT proteins were immunogenic, as they elicited antigen-specific cytotoxic T-lymphocyte responses in mice. These data support the utility of the COT approach to create novel and biologically active ancestral proteins as a starting point for studies of the structure, function, and biological fitness of highly variable genes, as well as for the rational design of globally relevant vaccine candidates.     

A new ATP-binding fold in actin, hexokinase and Hsc70

One of a cell biologist's favourite occupations is to discover the proteins that perform newly described functions in the cell. Very often lately, this has resulted in the identification of protein families whose related amino acid sequences reflect similar functions, but can proteins with totally unrelated sequences have similar structures and functions? In this review, Ken Holmes, Chris Sander and Alfonso Valencia describe the structural similarities between three well-known proteins that have no readily detectable primary sequence similarities but for which X-ray crystallography has revealed very similar structures. A comparison of their structures provides insights into their common mechanisms of action and into protein evolution, and has been used to detect related proteins in sequence data bases.     

Role of low-complexity sequences in the formation of novel protein coding sequences

Low-complexity sequences are extremely abundant in eukaryotic proteins for reasons that remain unclear. One hypothesis is that they contribute to the formation of novel coding sequences, facilitating the generation of novel protein functions. Here, we test this hypothesis by examining the content of low-complexity sequences in proteins of different age. We show that recently emerged proteins contain more low-complexity sequences than older proteins and that these sequences often form functional domains. These data are consistent with the idea that low-complexity sequences may play a key role in the emergence of novel genes.