Discovery of Novel Gain-of-Function Mutations Guided by Structure-Based Deep Learning

The impact of metagenomic analysis on the discovery of novel endolysins

Metagenomics has revolutionized enzyme discovery by enabling the study of genetic material directly from environmental samples, bypassing the need for microbial cultivation. This approach is particularly effective for identifying novel endolysins, phage-derived enzymes with antibacterial properties suited for therapeutic and industrial applications. Diverse ecosystems, such as biofilms, human microbiome, hot springs, and geothermal areas, serve as rich reservoirs for endolysins with traits like thermostability, broad-spectrum activity, specificity and resistance to harsh conditions. Functional metagenomics, complemented by bioinformatics, enables the discovery and annotation of previously uncharacterized endolysins. Examples of endolysins discovered from metagenomics analysis are discussed. Despite the challenges of analyzing complex microbial ecosystems and isolating target genes, metagenomics holds immense potential for uncovering innovative endolysins, paving the way for developing new biotechnological applications. KEY POINTS: • Endolysins offer antibacterial potential for therapeutic and industrial use. • Metagenomics enables discovery of novel endolysins from diverse ecosystems. • Advances in tools and methods have accelerated novel endolysins discovery.

Engineering Dehalogenase Enzymes Using Variational Autoencoder-Generated Latent Spaces and Microfluidics

Enzymes play a crucial role in sustainable industrial applications, with their optimization posing a formidable challenge due to the intricate interplay among residues. Computational methodologies predominantly rely on evolutionary insights of homologous sequences. However, deciphering the evolutionary variability and complex dependencies among residues presents substantial hurdles. Here, we present a new machine-learning method based on variational autoencoders and evolutionary sampling strategy to address those limitations. We customized our method to generate novel sequences of model enzymes, haloalkane dehalogenases. Three design-build-test cycles improved the solubility of variants from 11% to 75%. Thorough experimental validation including the microfluidic device MicroPEX resulted in 20 multiple-point variants. Nine of them, sharing as little as 67% sequence similarity with the template, showed a melting temperature increase of up to 9 °C and an average improvement of 3 °C. The most stable variant demonstrated a 3.5-fold increase in activity compared to the template. High-quality experimental data collected with 20 variants represent a valuable data set for the critical validation of novel protein design approaches. Python scripts, jupyter notebooks, and data sets are available on GitHub (https://github.com/loschmidt/vae-dehalogenases), and interactive calculations will be possible via https://loschmidt.chemi.muni.cz/fireprotasr/.

Enhancing recombinant growth factor and serum protein production for cultivated meat manufacturing

The commercial growth factors (GFs) and serum proteins (SPs) contribute to the high cost associated with the serum-free media for cultivated meat production. Producing recombinant GFs and SPs in scale from microbial cell factories can reduce the cost of culture media. Escherichia coli is a frequently employed host in the expression recombinant GFs and SPs. This review explores critical strategies for cost reduction in GFs and SPs production, focusing on yield enhancement, product improvement, purification innovation, and process innovation. Firstly, the review discusses the use of fusion tags to increase the solubility and yield of GFs & SPs, highlighting various studies that have successfully employed these tags for yield enhancement. We then explore how tagging strategies can streamline and economize the purification process, further reducing production costs. Additionally, we address the challenge of low half-life in GFs and SPs and propose potential strategies that can enhance their stability. Furthermore, improvements in the E. coli chassis and cell engineering strategies are also described, with an emphasis on the key areas that can improve yield and identify areas for cost minimization. Finally, we discuss key bioprocessing areas which can facilitate easier scale-up, enhance yield, titer, and productivity, and ultimately lower long-term production costs. It is crucial to recognize that not all suggested approaches can be applied simultaneously, as their relevance varies with different GFs and SPs. However, integrating of multiple strategies is anticipated to yield a cumulative effect, significantly reducing production costs. This collective effort is expected to substantially decrease the price of cultivated meat, contributing to the broader goal of developing sustainable and affordable meat.

Self-supervised machine learning methods for protein design improve sampling but not the identification of high-fitness variants

Machine learning (ML) is changing the world of computational protein design, with data-driven methods surpassing biophysical-based methods in experimental success. However, they are most often reported as case studies, lack integration and standardization, and are therefore hard to objectively compare. In this study, we established a streamlined and diverse toolbox for methods that predict amino acid probabilities inside the Rosetta software framework that allows for the side-by-side comparison of these models. Subsequently, existing protein fitness landscapes were used to benchmark novel ML methods in realistic protein design settings. We focused on the traditional problems of protein design: sampling and scoring. A major finding of our study is that ML approaches are better at purging the sampling space from deleterious mutations. Nevertheless, scoring resulting mutations without model fine-tuning showed no clear improvement over scoring with Rosetta. We conclude that ML now complements, rather than replaces, biophysical methods in protein design.

Tailoring industrial enzymes for thermostability and activity evolution by the machine learning-based iCASE strategy

The pursuit of obtaining enzymes with high activity and stability remains a grail in enzyme evolution due to the stability-activity trade-off. Here, we develop an isothermal compressibility-assisted dynamic squeezing index perturbation engineering (iCASE) strategy to construct hierarchical modular networks for enzymes of varying complexity. Molecular mechanism analysis elucidates that the peak of adaptive evolution is reached through a structural response mechanism among variants. Furthermore, this dynamic response predictive model using structure-based supervised machine learning is established to predict enzyme function and fitness, demonstrating robust performance across different datasets and reliable prediction for epistasis. The universality of the iCASE strategy is validated by four sorts of enzymes with different structures and catalytic types. This machine learning-based iCASE strategy provides guidance for future research on the fitness evolution of enzymes.

Computational Stabilization of a Non-Heme Iron Enzyme Enables Efficient Evolution of New Function

Deep learning tools for enzyme design are rapidly emerging, and there is a critical need to evaluate their effectiveness in engineering workflows. Here we show that the deep learning-based tool ProteinMPNN can be used to redesign Fe(II)/αKG superfamily enzymes for greater stability, solubility, and expression while retaining both native activity and industrially relevant non-native functions. This superfamily has diverse catalytic functions and could provide a rich new source of biocatalysts for synthesis and industrial processes. Through systematic comparisons of directed evolution trajectories for a non-native, remote C(sp3)-H hydroxylation reaction, we demonstrate that the stabilized redesign can be evolved more efficiently than the wild-type enzyme. After three rounds of directed evolution, we obtained a 6-fold activity increase from the wild-type parent and an 80-fold increase from the stabilized variant. To generate the initial stabilized variant, we identified multiple structural and sequence constraints to preserve catalytic function. We applied these criteria to produce stabilized, catalytically active variants of a second Fe(II)/αKG enzyme, suggesting that the approach is generalizable to additional members of the Fe(II)/αKG superfamily. ProteinMPNN is user-friendly and widely accessible, and our results provide a framework for the routine implementation of deep learning-based protein stabilization tools in directed evolution workflows for novel biocatalysts.

Machine Learning Guided Rational Design of a Non-Heme Iron-Based Lysine Dioxygenase Improves its Total Turnover Number

Highly selective C-H functionalization remains an ongoing challenge in organic synthetic methodologies. Biocatalysts are robust tools for achieving these difficult chemical transformations. Biocatalyst engineering has often required directed evolution or structure-based rational design campaigns to improve their activities. In recent years, machine learning has been integrated into these workflows to improve the discovery of beneficial enzyme variants. In this work, we combine a structure-based self-supervised machine learning framework, MutComputeX, with classical molecular dynamics simulations to down select mutations for rational design of a non-heme iron-dependent lysine dioxygenase, LDO. This approach consistently resulted in functional LDO mutants and circumvents the need for extensive study of mutational activity before-hand. Our rationally designed single mutants purified with up to 2-fold higher expression yields than WT and displayed higher total turnover numbers (TTN). Combining five such single mutations into a pentamutant variant, LPNYI LDO, leads to a 40 % improvement in the TTN (218±3) as compared to WT LDO (TTN=160±2). Overall, this work offers a low-barrier approach for those seeking to synergize machine learning algorithms with pre-existing protein engineering strategies.

Molecular dynamics and machine learning stratify motion-dependent activity profiles of S-layer destabilizing nanobodies

Nanobody (Nb)-induced disassembly of surface array protein (Sap) S-layers, a two-dimensional paracrystalline protein lattice from Bacillus anthracis, has been presented as a therapeutic intervention for lethal anthrax infections. However, only a subset of existing Nbs with affinity to Sap exhibit depolymerization activity, suggesting that affinity and epitope recognition are not enough to explain inhibitory activity. In this study, we performed all-atom molecular dynamics simulations of each Nb bound to the Sap binding site and trained a collection of machine learning classifiers to predict whether each Nb induces depolymerization. We used feature importance analysis to filter out unnecessary features and engineered remaining features to regularize the feature landscape and encourage learning of the depolymerization mechanism. We find that, while not enforced in training, a gradient-boosting decision tree is able to reproduce the experimental activities of inhibitory Nbs while maintaining high classification accuracy, whereas neural networks were only able to discriminate between classes. Further feature analysis revealed that inhibitory Nbs restrain Sap motions toward an inhibitory conformational state described by domain-domain clamping and induced twisting of domains normal to the lattice plane. We believe these motions drive Sap lattice depolymerization and can be used as design targets for improved Sap-inhibitory Nbs. Finally, we expect our method of study to apply to S-layers that serve as virulence factors in other pathogens, paving the way forward for Nb therapeutics that target depolymerization mechanisms.

Protein representations: Encoding biological information for machine learning in biocatalysis

Enzymes offer a more environmentally friendly and low-impact solution to conventional chemistry, but they often require additional engineering for their application in industrial settings, an endeavour that is challenging and laborious. To address this issue, the power of machine learning can be harnessed to produce predictive models that enable the in silico study and engineering of improved enzymatic properties. Such machine learning models, however, require the conversion of the complex biological information to a numerical input, also called protein representations. These inputs demand special attention to ensure the training of accurate and precise models, and, in this review, we therefore examine the critical step of encoding protein information to numeric representations for use in machine learning. We selected the most important approaches for encoding the three distinct biological protein representations - primary sequence, 3D structure, and dynamics - to explore their requirements for employment and inductive biases. Combined representations of proteins and substrates are also introduced as emergent tools in biocatalysis. We propose the division of fixed representations, a collection of rule-based encoding strategies, and learned representations extracted from the latent spaces of large neural networks. To select the most suitable protein representation, we propose two main factors to consider. The first one is the model setup, which is influenced by the size of the training dataset and the choice of architecture. The second factor is the model objectives such as consideration about the assayed property, the difference between wild-type models and mutant predictors, and requirements for explainability. This review is aimed at serving as a source of information and guidance for properly representing enzymes in future machine learning models for biocatalysis.

Structure and Energetics of PET-Hydrolyzing Enzyme Complexes: A Systematic Comparison from Molecular Dynamics Simulations

Discovered in 2016, the enzyme PETase, secreted by bacterial Ideonella Sakaiensis 201-F6, has an excellent hydrolytic activity toward poly(ethylene terephthalate) (PET) at room temperature, while it decreases at higher temperatures due to the low thermostability. Many variants have been engineered to overcome this limitation, which hinders industrial application. In this work, we systematically compare PETase wild-type (WT) and four mutants (DuraPETase, ThermoPETase, FastPETase, and HotPETase) using standard molecular dynamics (MD) simulations and unbinding free energy calculations. In particular, we analyze the enzymes' structural characteristics and binding to a tetrameric PET chain (PET4) under two temperature conditions: T1─300 K and T2─350 K. Our results indicate that (i) PET4 forms stable complexes with the five enzymes at room temperature (∼300 K) and (ii) most of the interactions are localized close to the active site of the protein, where the W185 and Y87 residues interact with the aromatic rings of the substrate. Specifically, (iii) the W185 side-chain explores different conformations in each variant (a phenomenon known in the literature as "W185 wobbling"). This suggests that the binding pocket retains structural plasticity and flexibility among the variants, facilitating substrate recognition and localization events at moderate temperatures. Moreover, (iv) PET4 establishes aromatic interactions with the catalytic H237 residue, stabilizing the catalytic triad composed of residues S160-H237-D206, and helping the system achieve an effective configuration for the hydrolysis reaction. Conversely, (v) the binding affinity decreases at a higher temperature (∼350 K), retaining moderate interactions only for HotPETase. Finally, (vi) MD simulations of complexes formed with poly(ethylene-2,5-furan dicarboxylate) (PEF) show no persistent interactions, suggesting that these enzymes are not yet optimized for binding this alternative semiaromatic plastic polymer. Our study offers valuable insights into the structural stability of these enzymes and the molecular determinants driving PET binding onto their surfaces, sheds light on the mechanistic steps that precede the onset of hydrolysis, and provides a foundation for future enzyme optimization.

Click, Compute, Create: A Review of Web-based Tools for Enzyme Engineering

Enzyme engineering, though pivotal across various biotechnological domains, is often plagued by its time-consuming and labor-intensive nature. This review aims to offer an overview of supportive in silico methodologies for this demanding endeavor. Starting from methods to predict protein structures, to classification of their activity and even the discovery of new enzymes we continue with describing tools used to increase thermostability and production yields of selected targets. Subsequently, we discuss computational methods to modulate both, the activity as well as selectivity of enzymes. Last, we present recent approaches based on cutting-edge machine learning methods to redesign enzymes. With exception of the last chapter, there is a strong focus on methods easily accessible via web-interfaces or simple Python-scripts, therefore readily useable for a diverse and broad community.

Machine learning-guided multi-site combinatorial mutagenesis enhances the thermostability of pectin lyase

Enhancing the thermostability of enzymes is crucial for industrial applications. Methods such as directed evolution are often limited by the huge sequence space and combinatorial explosion, making it difficult to obtain optimal mutants. In recent years, machine learning (ML)-guided protein engineering has become an attractive tool because of its ability to comprehensively explore the sequence space of enzymes and discover superior mutants. This study employed ML to perform combinatorial mutation design on the pectin lyase PMGL-Ba from Bacillus licheniformis, aiming to improve its thermostability. First, 18 single-point mutants with enhanced thermostability were identified through semi-rational design. Subsequently, the initial library containing a small number of low-order mutants was utilized to construct an ML model to explore the combinatorial sequence space (theoretically 196,608 mutants) of single-point mutants. The results showed that the ML-predicted second library was successfully enriched with highly thermostable combinatorial mutants. After one iteration of learning, the best-performing combinatorial mutant in the third library, P36, showed a 67-fold and 39-fold increase in half-life at 75 °C and 80 °C, respectively, as well as a 2.1-fold increase in activity. Structural analysis and molecular dynamics simulations provided insights into the improved performance of the engineered enzyme.

Hook loop dynamics engineering transcended the barrier of activity-stability trade-off and boosted the thermostability of enzymes

The improvement of enzyme thermostability often accompanies the decreased activity due to the loss of the key regions' flexibility. As a representative structure, unlocking the potential of loop dynamics will not only provide new ideas for stabilization strategies, but also help to deepen the understanding of the relationship between enzyme structural dynamics and function. In this study, a creative "hook loop dynamics engineering" (HLoD) strategy was successfully proposed for simultaneously improving the thermostability and maintaining activity of the model enzyme, Candida Antarctica lipase B. A small and smart mutant library involving five key residues located at the "hook loop" was meticulously identified and systematically investigated and thus yielded a five-point multiple mutant M1 (L147S/T244P/S250P/T256D/N292D), demonstrating a remarkable 7.0-fold increase in thermostability at 60 °C compared to the wild-type (WT). Furthermore, the activity of M1 remained comparable to that of WT, effectively transcending the barrier of activity-stability trade-off. Molecular dynamics simulations revealed that the precise regulation of hook loop dynamics via intermolecular interactions, such as salt bridges and hydrogen bonding, curbed the excessive flexibility of the pivotal regions α5 and α10 at high temperatures, thus driving the substantial enhancement of the thermostability of M1. Refining the dynamics of the flexible region via HLoD, which transcended the barrier of activity-stability trade-off, exhibited to be a robust and potentially universal strategy for designing enzymes with outstanding thermostability and activity.

On synergy between ultrahigh throughput screening and machine learning in biocatalyst engineering

Protein design and directed evolution have separately contributed enormously to protein engineering. Without being mutually exclusive, the former relies on computation from first principles, while the latter is a combinatorial approach based on chance. Advances in ultrahigh throughput (uHT) screening, next generation sequencing and machine learning may create alternative routes to engineered proteins, where functional information linked to specific sequences is interpreted and extrapolated in silico. In particular, the miniaturisation of functional tests in water-in-oil emulsion droplets with picoliter volumes and their rapid generation and analysis (>1 kHz) allows screening of >107-membered libraries in a day. Subsequently, decoding the selected clones by short or long-read sequencing methods leads to large sequence-function datasets that may allow extrapolation from experimental directed evolution to further improved mutants beyond the observed hits. In this work, we explore experimental strategies for how to draw up 'fitness landscapes' in sequence space with uHT droplet microfluidics, review the current state of AI/ML in enzyme engineering and discuss how uHT datasets may be combined with AI/ML to make meaningful predictions and accelerate biocatalyst engineering.

Advancing microbial production through artificial intelligence-aided biology

Microbial cell factories (MCFs) have been leveraged to construct sustainable platforms for value-added compound production. To optimize metabolism and reach optimal productivity, synthetic biology has developed various genetic devices to engineer microbial systems by gene editing, high-throughput protein engineering, and dynamic regulation. However, current synthetic biology methodologies still rely heavily on manual design, laborious testing, and exhaustive analysis. The emerging interdisciplinary field of artificial intelligence (AI) and biology has become pivotal in addressing the remaining challenges. AI-aided microbial production harnesses the power of processing, learning, and predicting vast amounts of biological data within seconds, providing outputs with high probability. With well-trained AI models, the conventional Design-Build-Test (DBT) cycle has been transformed into a multidimensional Design-Build-Test-Learn-Predict (DBTLP) workflow, leading to significantly improved operational efficiency and reduced labor consumption. Here, we comprehensively review the main components and recent advances in AI-aided microbial production, focusing on genome annotation, AI-aided protein engineering, artificial functional protein design, and AI-enabled pathway prediction. Finally, we discuss the challenges of integrating novel AI techniques into biology and propose the potential of large language models (LLMs) in advancing microbial production.

A modular toolkit for environmental Rhodococcus, Gordonia, and Nocardia enables complex metabolic manipulation

Soil-dwelling Actinomycetes are a diverse and ubiquitous component of the global microbiome but largely lack genetic tools comparable to those available in model species such as Escherichia coli or Pseudomonas putida, posing a fundamental barrier to their characterization and utilization as hosts for biotechnology. To address this, we have developed a modular plasmid assembly framework, along with a series of genetic control elements for the previously genetically intractable Gram-positive environmental isolate Rhodococcus ruber C208, and demonstrate conserved functionality in 11 additional environmental isolates of Rhodococcus, Nocardia, and Gordonia. This toolkit encompasses five Mycobacteriale origins of replication, five broad-host-range antibiotic resistance markers, transcriptional and translational control elements, fluorescent reporters, a tetracycline-inducible system, and a counter-selectable marker. We use this toolkit to interrogate the carotenoid biosynthesis pathway in Rhodococcus erythropolis N9T-4, a weakly carotenogenic environmental isolate and engineer higher pathway flux toward the keto-carotenoid canthaxanthin. This work establishes several new genetic tools for environmental Mycobacteriales and provides a synthetic biology framework to support the design of complex genetic circuits in these species.IMPORTANCESoil-dwelling Actinomycetes, particularly the Mycobacteriales, include both diverse new hosts for sustainable biomanufacturing and emerging opportunistic pathogens. Rhodococcus, Gordonia, and Nocardia are three abundant genera with particularly flexible metabolisms and untapped potential for natural product discovery. Among these, Rhodococcus ruber C208 was shown to degrade polyethylene; Gordonia paraffinivorans can assimilate carbon from solid hydrocarbons; and Nocardia neocaledoniensis (and many other Nocardia spp.) possesses dual isoprenoid biosynthesis pathways. Many species accumulate high levels of carotenoid pigments, indicative of highly active isoprenoid biosynthesis pathways which may be harnessed for fermentation of terpenes and other commodity isoprenoids. Modular genetic toolkits have proven valuable for both fundamental and applied research in model organisms, but such tools are lacking for most Actinomycetes. Our suite of genetic tools and DNA assembly framework were developed for broad functionality and to facilitate rapid prototyping of genetic constructs in these organisms.

EnzyACT: A Novel Deep Learning Method to Predict the Impacts of Single and Multiple Mutations on Enzyme Activity

Enzyme engineering involves the customization of enzymes by introducing mutations to expand the application scope of natural enzymes. One limitation of that is the complex interaction between two key properties, activity and stability, where the enhancement of one often leads to the reduction of the other, also called the trade-off mechanism. Although dozens of methods that predict the change of protein stability upon mutations have been developed, the prediction of the effect on activity is still in its early stage. Therefore, developing a fast and accurate method to predict the impact of the mutations on enzyme activity is helpful for enzyme design and understanding of the trade-off mechanism. Here, we introduce a novel approach, EnzyACT, a deep learning method that fuses graph technique and protein embedding to predict activity changes upon single or multiple mutations. Our model combines graph-based techniques and language models to predict the activity changes. Moreover, EnzyACT is trained on a new curated data set including both single- and multiple-point mutations. When benchmarked on multiple independent data sets, it shows uniform performance on problems affected by mutations. This work also provides insights into the impact of distant mutations within activity design, which could also be useful for predicting catalytic residues and developing improved enzyme-engineering strategies.

Navigating the landscape of enzyme design: from molecular simulations to machine learning

Global environmental issues and sustainable development call for new technologies for fine chemical synthesis and waste valorization. Biocatalysis has attracted great attention as the alternative to the traditional organic synthesis. However, it is challenging to navigate the vast sequence space to identify those proteins with admirable biocatalytic functions. The recent development of deep-learning based structure prediction methods such as AlphaFold2 reinforced by different computational simulations or multiscale calculations has largely expanded the 3D structure databases and enabled structure-based design. While structure-based approaches shed light on site-specific enzyme engineering, they are not suitable for large-scale screening of potential biocatalysts. Effective utilization of big data using machine learning techniques opens up a new era for accelerated predictions. Here, we review the approaches and applications of structure-based and machine-learning guided enzyme design. We also provide our view on the challenges and perspectives on effectively employing enzyme design approaches integrating traditional molecular simulations and machine learning, and the importance of database construction and algorithm development in attaining predictive ML models to explore the sequence fitness landscape for the design of admirable biocatalysts.

A synthetic cell-free pathway for biocatalytic upgrading of one-carbon substrates

Biotechnological processes hold tremendous potential for the efficient and sustainable conversion of one-carbon (C1) substrates into complex multi-carbon products. However, the development of robust and versatile biocatalytic systems for this purpose remains a significant challenge. In this study, we report a hybrid electrochemical-biochemical cell-free system for the conversion of C1 substrates into the universal biological building block acetyl-CoA. The synthetic reductive formate pathway (ReForm) consists of five core enzymes catalyzing non-natural reactions that were established through a cell-free enzyme engineering platform. We demonstrate that ReForm works in a plug-and-play manner to accept diverse C1 substrates including CO2 equivalents. We anticipate that ReForm will facilitate efforts to build and improve synthetic C1 utilization pathways for a formate-based bioeconomy.

Cell factories for biosynthesis of D-glucaric acid: a fusion of static and dynamic strategies

D-glucaric acid is an important organic acid with numerous applications in therapy, food, and materials, contributing significantly to its substantial market value. The biosynthesis of D-glucaric acid (GA) from renewable sources such as glucose has garnered significant attention due to its potential for sustainable and cost-effective production. This review summarizes the current understanding of the cell factories for GA production in different chassis strains, from static to dynamic control strategies for regulating their metabolic networks. We highlight recent advances in the optimization of D-glucaric acid biosynthesis, including metabolic dynamic control, alternative feedstocks, metabolic compartments, and so on. Additionally, we compare the differences between different chassis strains and discuss the challenges that each chassis strain must overcome to achieve highly efficient GA productions. In this review, the processes of engineering a desirable cell factory for highly efficient GA production are just like an epitome of metabolic engineering of strains for chemical biosynthesis, inferring general trends for industrial chassis strain developments.

Stability Oracle: a structure-based graph-transformer framework for identifying stabilizing mutations

Engineering stabilized proteins is a fundamental challenge in the development of industrial and pharmaceutical biotechnologies. We present Stability Oracle: a structure-based graph-transformer framework that achieves SOTA performance on accurately identifying thermodynamically stabilizing mutations. Our framework introduces several innovations to overcome well-known challenges in data scarcity and bias, generalization, and computation time, such as: Thermodynamic Permutations for data augmentation, structural amino acid embeddings to model a mutation with a single structure, a protein structure-specific attention-bias mechanism that makes transformers a viable alternative to graph neural networks. We provide training/test splits that mitigate data leakage and ensure proper model evaluation. Furthermore, to examine our data engineering contributions, we fine-tune ESM2 representations (Prostata-IFML) and achieve SOTA for sequence-based models. Notably, Stability Oracle outperforms Prostata-IFML even though it was pretrained on 2000X less proteins and has 548X less parameters. Our framework establishes a path for fine-tuning structure-based transformers to virtually any phenotype, a necessary task for accelerating the development of protein-based biotechnologies.

Machine learning guided rational design of a non-heme iron-based lysine dioxygenase improves its total turnover number

Highly selective C-H functionalization remains an ongoing challenge in organic synthetic methodologies. Biocatalysts are robust tools for achieving these difficult chemical transformations. Biocatalyst engineering has often required directed evolution or structure-based rational design campaigns to improve their activities. In recent years, machine learning has been integrated into these workflows to improve the discovery of beneficial enzyme variants. In this work, we combine a structure-based machine-learning algorithm with classical molecular dynamics simulations to down select mutations for rational design of a non-heme iron-dependent lysine dioxygenase, LDO. This approach consistently resulted in functional LDO mutants and circumvents the need for extensive study of mutational activity before-hand. Our rationally designed single mutants purified with up to 2-fold higher yields than WT and displayed higher total turnover numbers (TTN). Combining five such single mutations into a pentamutant variant, LPNYI LDO, leads to a 40% improvement in the TTN (218±3) as compared to WT LDO (TTN = 160±2). Overall, this work offers a low-barrier approach for those seeking to synergize machine learning algorithms with pre-existing protein engineering strategies.

GMMA Can Stabilize Proteins Across Different Functional Constraints

Stabilizing proteins without otherwise hampering their function is a central task in protein engineering and design. PYR1 is a plant hormone receptor that has been engineered to bind diverse small molecule ligands. We sought a set of generalized mutations that would provide stability without affecting functionality for PYR1 variants with diverse ligand-binding capabilities. To do this we used a global multi-mutant analysis (GMMA) approach, which can identify substitutions that have stabilizing effects and do not lower function. GMMA has the added benefit of finding substitutions that are stabilizing in different sequence contexts and we hypothesized that applying GMMA to PYR1 with different functionalities would identify this set of generalized mutations. Indeed, conducting FACS and deep sequencing of libraries for PYR1 variants with two different functionalities and applying a GMMA analysis identified 5 substitutions that, when inserted into four PYR1 variants that each bind a unique ligand, provided an increase of 2-6 °C in thermal inactivation temperature and no decrease in functionality.

Enhancing PET Degrading Enzymes: A Combinatory Approach

Plastic waste has become a substantial environmental issue. A potential strategy to mitigate this problem is to use enzymatic hydrolysis of plastics to depolymerize post-consumer waste and allow it to be reused. Over the last few decades, the use of enzymatic PET-degrading enzymes has shown promise as a great solution for creating a circular plastic waste economy. PsPETase from Piscinibacter sakaiensis has been identified as an enzyme with tremendous potential for such applications. But to improve its efficiency, enzyme engineering has been applied aiming at enhancing its thermal stability, enzymatic activity, and ease of production. Here, we combine different strategies such as structure-based rational design, ancestral sequence reconstruction and machine learning to engineer a more highly active Combi-PETase variant with a melting temperature of 70 °C and optimal performance at 60 °C. Furthermore, this study demonstrates that these approaches, commonly used in other works of enzyme engineering, are most effective when utilized in combination, enabling the improvement of enzymes for industrial applications.

Reconstructing dynamics correlation network to simultaneously improve activity and stability of 2,3-butanediol dehydrogenase by design of distal interchain disulfide bonds

The complete enzyme catalytic cycle includes substrate binding, chemical reaction and product release, in which different dynamic conformations are adopted. Due to the complex relationship among enzyme activity, stability and dynamics, the directed evolution of enzymes for improved activity or stability commonly leads to a trade-off in stability or activity. It hence remains a challenge to engineer an enzyme to have both enhanced activity and stability. Here, we have attempted to reconstruct the dynamics correlation network involved with active center to improve both activity and stability of a 2,3-butanediol dehydrogenase (2,3-BDH) by introducing inter-chain disulfide bonds. A computational strategy was first applied to evaluate the effect of introducing inter-chain disulfide bond on activity and stability of three 2,3-BDHs, and the N258C mutation of 2,3-BDH from Corynebacterium glutamicum (CgBDH) was proved to be effective in improving both activity and stability. In the results, CgBDH-N258C showed a different unfolding curve from the wild type, with two melting temperatures (Tm) of 68.3 °C and 50.8 °C, 19.7 °C and 2 °C higher than 48.6 °C of the wild type. Its half-life was also improved by 14.8-fold compared to the wild type. Catalytic efficiency (kcat/Km) of the mutant was increased by 7.9-fold toward native substrate diacetyl and 8.8-fold toward non-native substrate 2,5-hexanedione compared to the wild type. Molecular dynamics simulations revealed that an interaction network formed by Cys258, Arg162, Ala144 and the catalytic residues was reconstructed in the mutant and the dynamics change caused by the disulfide bond could be propagated through the interactions network. This improved the enzyme stability and activity by decreasing the flexibility and locking more "reactive" pose, respectively. Further construction of mutations including A144G showing a 44-fold improvement in catalytic efficiency toward meso-2,3-BD confirmed the role of modifying dynamics correlation network in tunning enzyme activity and selectivity. This study provided important insights into the relationship among dynamics, enzyme catalysis and stability, and will be useful in the designing new enzymes with co-evolution of stability, activity and selectivity.

Enhancing missense variant pathogenicity prediction with protein language models using VariPred

Computational approaches for predicting the pathogenicity of genetic variants have advanced in recent years. These methods enable researchers to determine the possible clinical impact of rare and novel variants. Historically these prediction methods used hand-crafted features based on structural, evolutionary, or physiochemical properties of the variant. In this study we propose a novel framework that leverages the power of pre-trained protein language models to predict variant pathogenicity. We show that our approach VariPred (Variant impact Predictor) outperforms current state-of-the-art methods by using an end-to-end model that only requires the protein sequence as input. Using one of the best-performing protein language models (ESM-1b), we establish a robust classifier that requires no calculation of structural features or multiple sequence alignments. We compare the performance of VariPred with other representative models including 3Cnet, Polyphen-2, REVEL, MetaLR, FATHMM and ESM variant. VariPred performs as well as, or in most cases better than these other predictors using six variant impact prediction benchmarks despite requiring only sequence data and no pre-processing of the data.

Asymmetric Synthesis of α-Chloroamides via Photoenzymatic Hydroalkylation of Olefins

Photoenzymatic intermolecular hydroalkylations of olefins are highly enantioselective for chiral centers formed during radical termination but poorly selective for centers set in the C-C bond-forming event. Here, we report the evolution of a flavin-dependent "ene"-reductase to catalyze the coupling of α,α-dichloroamides with alkenes to afford α-chloroamides in good yield with excellent chemo- and stereoselectivity. These products can serve as linchpins in the synthesis of pharmaceutically valuable motifs. Mechanistic studies indicate that radical formation occurs by exciting a charge-transfer complex templated by the protein. Precise control over the orientation of molecules within the charge-transfer complex potentially accounts for the observed stereoselectivity. The work expands the types of motifs that can be prepared using photoenzymatic catalysis.

Biosensor and machine learning-aided engineering of an amaryllidaceae enzyme

A major challenge to achieving industry-scale biomanufacturing of therapeutic alkaloids is the slow process of biocatalyst engineering. Amaryllidaceae alkaloids, such as the Alzheimer's medication galantamine, are complex plant secondary metabolites with recognized therapeutic value. Due to their difficult synthesis they are regularly sourced by extraction and purification from the low-yielding daffodil Narcissus pseudonarcissus. Here, we propose an efficient biosensor-machine learning technology stack for biocatalyst development, which we apply to engineer an Amaryllidaceae enzyme in Escherichia coli. Directed evolution is used to develop a highly sensitive (EC50 = 20 μM) and specific biosensor for the key Amaryllidaceae alkaloid branchpoint 4'-O-methylnorbelladine. A structure-based residual neural network (MutComputeX) is subsequently developed and used to generate activity-enriched variants of a plant methyltransferase, which are rapidly screened with the biosensor. Functional enzyme variants are identified that yield a 60% improvement in product titer, 2-fold higher catalytic activity, and 3-fold lower off-product regioisomer formation. A solved crystal structure elucidates the mechanism behind key beneficial mutations.

Machine Learning and Deep Learning in Synthetic Biology: Key Architectures, Applications, and Challenges

Machine learning (ML), particularly deep learning (DL), has made rapid and substantial progress in synthetic biology in recent years. Biotechnological applications of biosystems, including pathways, enzymes, and whole cells, are being probed frequently with time. The intricacy and interconnectedness of biosystems make it challenging to design them with the desired properties. ML and DL have a synergy with synthetic biology. Synthetic biology can be employed to produce large data sets for training models (for instance, by utilizing DNA synthesis), and ML/DL models can be employed to inform design (for example, by generating new parts or advising unrivaled experiments to perform). This potential has recently been brought to light by research at the intersection of engineering biology and ML/DL through achievements like the design of novel biological components, best experimental design, automated analysis of microscopy data, protein structure prediction, and biomolecular implementations of ANNs (Artificial Neural Networks). I have divided this review into three sections. In the first section, I describe predictive potential and basics of ML along with myriad applications in synthetic biology, especially in engineering cells, activity of proteins, and metabolic pathways. In the second section, I describe fundamental DL architectures and their applications in synthetic biology. Finally, I describe different challenges causing hurdles in the progress of ML/DL and synthetic biology along with their solutions.

Opportunities and Challenges for Machine Learning-Assisted Enzyme Engineering

Enzymes can be engineered at the level of their amino acid sequences to optimize key properties such as expression, stability, substrate range, and catalytic efficiency-or even to unlock new catalytic activities not found in nature. Because the search space of possible proteins is vast, enzyme engineering usually involves discovering an enzyme starting point that has some level of the desired activity followed by directed evolution to improve its "fitness" for a desired application. Recently, machine learning (ML) has emerged as a powerful tool to complement this empirical process. ML models can contribute to (1) starting point discovery by functional annotation of known protein sequences or generating novel protein sequences with desired functions and (2) navigating protein fitness landscapes for fitness optimization by learning mappings between protein sequences and their associated fitness values. In this Outlook, we explain how ML complements enzyme engineering and discuss its future potential to unlock improved engineering outcomes.

Learning the shape of protein microenvironments with a holographic convolutional neural network

Proteins play a central role in biology from immune recognition to brain activity. While major advances in machine learning have improved our ability to predict protein structure from sequence, determining protein function from its sequence or structure remains a major challenge. Here, we introduce holographic convolutional neural network (H-CNN) for proteins, which is a physically motivated machine learning approach to model amino acid preferences in protein structures. H-CNN reflects physical interactions in a protein structure and recapitulates the functional information stored in evolutionary data. H-CNN accurately predicts the impact of mutations on protein stability and binding of protein complexes. Our interpretable computational model for protein structure-function maps could guide design of novel proteins with desired function.

Data-Driven Protein Engineering for Improving Catalytic Activity and Selectivity

Protein engineering is essential for altering the substrate scope, catalytic activity and selectivity of enzymes for applications in biocatalysis. However, traditional approaches, such as directed evolution and rational design, encounter the challenge in dealing with the experimental screening process of a large protein mutation space. Machine learning methods allow the approximation of protein fitness landscapes and the identification of catalytic patterns using limited experimental data, thus providing a new avenue to guide protein engineering campaigns. In this concept article, we review machine learning models that have been developed to assess enzyme-substrate-catalysis performance relationships aiming to improve enzymes through data-driven protein engineering. Furthermore, we prospect the future development of this field to provide additional strategies and tools for achieving desired activities and selectivities.

Improving Protein Expression, Stability, and Function with ProteinMPNN

Natural proteins are highly optimized for function but are often difficult to produce at a scale suitable for biotechnological applications due to poor expression in heterologous systems, limited solubility, and sensitivity to temperature. Thus, a general method that improves the physical properties of native proteins while maintaining function could have wide utility for protein-based technologies. Here, we show that the deep neural network ProteinMPNN, together with evolutionary and structural information, provides a route to increasing protein expression, stability, and function. For both myoglobin and tobacco etch virus (TEV) protease, we generated designs with improved expression, elevated melting temperatures, and improved function. For TEV protease, we identified multiple designs with improved catalytic activity as compared to the parent sequence and previously reported TEV variants. Our approach should be broadly useful for improving the expression, stability, and function of biotechnologically important proteins.

Mammalian Antigen Display for Pandemic Countermeasures

Pandemic countermeasures require the rapid design of antigens for vaccines, profiling patient antibody responses, assessing antigen structure-function landscapes, and the surveillance of emerging viral lineages. Cell surface display of a viral antigen or its subdomains can facilitate these goals by coupling the phenotypes of protein variants to their DNA sequence. Screening surface-displayed proteins via flow cytometry also eliminates time-consuming protein purification steps. Prior approaches have primarily relied on yeast as a display chassis. However, yeast often cannot express large viral glycoproteins, requiring their truncation into subdomains. Here, we describe a method to design and express antigens on the surface of mammalian HEK293T cells. We discuss three use cases, including screening of stabilizing mutations, deep mutational scanning, and epitope mapping. The mammalian antigen display platform described herein will accelerate ongoing and future pandemic countermeasures.

Engineering Enzymes for Environmental Sustainability

The development and implementation of sustainable catalytic technologies is key to delivering our net-zero targets. Here we review how engineered enzymes, with a focus on those developed using directed evolution, can be deployed to improve the sustainability of numerous processes and help to conserve our environment. Efficient and robust biocatalysts have been engineered to capture carbon dioxide (CO2) and have been embedded into new efficient metabolic CO2 fixation pathways. Enzymes have been refined for bioremediation, enhancing their ability to degrade toxic and harmful pollutants. Biocatalytic recycling is gaining momentum, with engineered cutinases and PETases developed for the depolymerization of the abundant plastic, polyethylene terephthalate (PET). Finally, biocatalytic approaches for accessing petroleum-based feedstocks and chemicals are expanding, using optimized enzymes to convert plant biomass into biofuels or other high value products. Through these examples, we hope to illustrate how enzyme engineering and biocatalysis can contribute to the development of cleaner and more efficient chemical industry.

Engineering Enzymes for Environmental Sustainability

The development and implementation of sustainable catalytic technologies is key to delivering our net-zero targets. Here we review how engineered enzymes, with a focus on those developed using directed evolution, can be deployed to improve the sustainability of numerous processes and help to conserve our environment. Efficient and robust biocatalysts have been engineered to capture carbon dioxide (CO2 ) and have been embedded into new efficient metabolic CO2 fixation pathways. Enzymes have been refined for bioremediation, enhancing their ability to degrade toxic and harmful pollutants. Biocatalytic recycling is gaining momentum, with engineered cutinases and PETases developed for the depolymerization of the abundant plastic, polyethylene terephthalate (PET). Finally, biocatalytic approaches for accessing petroleum-based feedstocks and chemicals are expanding, using optimized enzymes to convert plant biomass into biofuels or other high value products. Through these examples, we hope to illustrate how enzyme engineering and biocatalysis can contribute to the development of cleaner and more efficient chemical industry.

Modifying the amino acids in conformational motion pathway of the α-amylase of Geobacillus stearothermophilus improved its activity and stability

Amino acids along the conformational motion pathway of the enzyme molecule correlated to its flexibility and rigidity. To enhance the enzyme activity and thermal stability, the motion pathway of Geobacillus stearothermophilus α-amylase has been identified and molecularly modified by using the neural relational inference model and deep learning tool. The significant differences in substrate specificity, enzymatic kinetics, optimal temperature, and thermal stability were observed among the mutants with modified amino acids along the pathway. Mutants especially the P44E demonstrated enhanced hydrolytic activity and catalytic efficiency (kcat/KM) than the wild-type enzyme to 95.0% and 93.8% respectively, with the optimum temperature increased to 90°C. This mutation from proline to glutamic acid has increased the number and the radius of the bottleneck of the channels, which might facilitate transporting large starch substrates into the enzyme. The mutation could also optimize the hydrogen bonding network of the catalytic center, and diminish the spatial hindering to the substrate entry and exit from the catalytic center.

Module-Based Polyketide Synthase Engineering for de Novo Polyketide Biosynthesis

Polyketide retrobiosynthesis, where the biosynthetic pathway of a given polyketide can be reversibly engineered due to the colinearity of the polyketide synthase (PKS) structure and function, has the potential to produce millions of organic molecules. Mixing and matching modules from natural PKSs is one of the routes to produce many of these molecules. Evolutionary analysis of PKSs suggests that traditionally used module boundaries may not lead to the most productive hybrid PKSs and that new boundaries around and within the ketosynthase domain may be more active when constructing hybrid PKSs. As this is still a nascent area of research, the generality of these design principles based on existing engineering efforts remains inconclusive. Recent advances in structural modeling and synthetic biology present an opportunity to accelerate PKS engineering by re-evaluating insights gained from previous engineering efforts with cutting edge tools.

Mining cancer genomes for change-of-metabolic-function mutations

Enzymes with novel functions are needed to enable new organic synthesis techniques. Drawing inspiration from gain-of-function cancer mutations that functionally alter proteins and affect cellular metabolism, we developed METIS (Mutated Enzymes from Tumors In silico Screen). METIS identifies metabolism-altering cancer mutations using mutation recurrence rates and protein structure. We used METIS to screen 298,517 cancer mutations and identify 48 candidate mutations, including those previously identified to alter enzymatic function. Unbiased metabolomic profiling of cells exogenously expressing a candidate mutant (OGDHLp.A400T) supports an altered phenotype that boosts in vitro production of xanthosine, a pharmacologically useful chemical that is currently produced using unsustainable, water-intensive methods. We then applied METIS to 49 million cancer mutations, yielding a refined set of candidates that may impart novel enzymatic functions or contribute to tumor progression. Thus, METIS can be used to identify and catalog potentially-useful cancer mutations for green chemistry and therapeutic applications.

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.

Discovering pathways through ribozyme fitness landscapes using information theoretic quantification of epistasis

The identification of catalytic RNAs is typically achieved through primarily experimental means. However, only a small fraction of sequence space can be analyzed even with high-throughput techniques. Methods to extrapolate from a limited data set to predict additional ribozyme sequences, particularly in a human-interpretable fashion, could be useful both for designing new functional RNAs and for generating greater understanding about a ribozyme fitness landscape. Using information theory, we express the effects of epistasis (i.e., deviations from additivity) on a ribozyme. This representation was incorporated into a simple model of the epistatic fitness landscape, which identified potentially exploitable combinations of mutations. We used this model to theoretically predict mutants of high activity for a self-aminoacylating ribozyme, identifying potentially active triple and quadruple mutants beyond the experimental data set of single and double mutants. The predictions were validated experimentally, with nine out of nine sequences being accurately predicted to have high activity. This set of sequences included mutants that form a previously unknown evolutionary "bridge" between two ribozyme families that share a common motif. Individual steps in the method could be examined, understood, and guided by a human, combining interpretability and performance in a simple model to predict ribozyme sequences by extrapolation.

Advancing Enzyme's Stability and Catalytic Efficiency through Synergy of Force-Field Calculations, Evolutionary Analysis, and Machine Learning

Thermostability is an essential requirement for the use of enzymes in the bioindustry. Here, we compare different protein stabilization strategies using a challenging target, a stable haloalkane dehalogenase DhaA115. We observe better performance of automated stabilization platforms FireProt and PROSS in designing multiple-point mutations over the introduction of disulfide bonds and strengthening the intra- and the inter-domain contacts by in silico saturation mutagenesis. We reveal that the performance of automated stabilization platforms was still compromised due to the introduction of some destabilizing mutations. Notably, we show that their prediction accuracy can be improved by applying manual curation or machine learning for the removal of potentially destabilizing mutations, yielding highly stable haloalkane dehalogenases with enhanced catalytic properties. A comparison of crystallographic structures revealed that current stabilization rounds were not accompanied by large backbone re-arrangements previously observed during the engineering stability of DhaA115. Stabilization was achieved by improving local contacts including protein-water interactions. Our study provides guidance for further improvement of automated structure-based computational tools for protein stabilization.

Design of SC PEP with enhanced stability against pepsin digestion and increased activity by machine learning and structural parameters modeling

Prolyl endopeptidases from Sphingomonas capsulata (SC PEP) has attracted much attention as promising oral therapy candidate for celiac sprue, however, its low stability in the gastric environment leads to unsatisfactory clinical results. Therefore, improving its stability against pepsin digestion at low pH is crucial for clinical applications, but challenging. In this study, machine learning and physical parameter model were combined to design SC PEP mutants. After iterations, 20 mutants had higher hydrolysis activity in stomach environment, which was up to 14.1-fold compared with wild-type SC PEP. Mutant M24 involving stable and active mutations and pegylated M24 (M24-PEG) had higher activity of hydrolyzing immunogen in bread than wild-type SC PEP in vitro and in vivo, and residual immunogens in simulated gastric environment were only 1/8 and 1/10 of that in the wild-type SC PEP group. The total residual immunogens in the gastrointestinal tract of mice in the M24 and M24-PEG groups were <20 ppm, reaching the standard of non-toxic food. Our results indicate that the combination of M24 (or M24-PEG) with EP-B2 may be a promising candidate for celiac disease, and the strategies developed in this study provide a paradigm for the design of SC PEP stability mutants.

Current advances in the structural biology and molecular engineering of PETase

Poly(ethylene terephthalate) (PET) is a highly useful synthetic polyester plastic that is widely used in daily life. However, the increase in postconsumer PET as plastic waste that is recalcitrant to biodegradation in landfills and the natural environment has raised worldwide concern. Currently, traditional PET recycling processes with thermomechanical or chemical methods also result in the deterioration of the mechanical properties of PET. Therefore, it is urgent to develop more efficient and green strategies to address this problem. Recently, a novel mesophilic PET-degrading enzyme (IsPETase) from Ideonella sakaiensis was found to streamline PET biodegradation at 30°C, albeit with a lower PET-degrading activity than chitinase or chitinase-like PET-degrading enzymes. Consequently, the molecular engineering of more efficient PETases is still required for further industrial applications. This review details current knowledge on IsPETase, MHETase, and IsPETase-like hydrolases, including the structures, ligand‒protein interactions, and rational protein engineering for improved PET-degrading performance. In particular, applications of the engineered catalysts are highlighted, including metabolic engineering of the cell factories, enzyme immobilization or cell surface display. The information is expected to provide novel insights for the biodegradation of complex polymers.

Engineering the catalytic activity of an Antarctic PET-degrading enzyme by loop exchange

Several hydrolases have been described to degrade polyethylene terephthalate (PET) at moderate temperatures ranging from 25°C to 40°C. These mesophilic PET hydrolases (PETases) are less efficient in degrading this plastic polymer than their thermophilic homologs and have, therefore, been the subject of many protein engineering campaigns. However, enhancing their enzymatic activity through rational design or directed evolution poses a formidable challenge due to the need for exploring a large number of mutations. Additionally, evaluating the improvements in both activity and stability requires screening numerous variants, either individually or using high-throughput screening methods. Here, we utilize instead the design of chimeras as a protein engineering strategy to increase the activity and stability of Mors1, an Antarctic PETase active at 25°C. First, we obtained the crystal structure of Mors1 at 1.6 Å resolution, which we used as a scaffold for structure- and sequence-based chimeric design. Then, we designed a Mors1 chimera via loop exchange of a highly divergent active site loop from the thermophilic leaf-branch compost cutinase (LCC) into the equivalent region in Mors1. After restitution of an active site disulfide bond into this chimera, the enzyme exhibited a shift in optimal temperature for activity to 45°C and an increase in fivefold in PET hydrolysis when compared with wild-type Mors1 at 25°C. Our results serve as a proof of concept of the utility of chimeric design to further improve the activity and stability of PETases active at moderate temperatures.

Two sequence- and two structure-based ML models have learned different aspects of protein biochemistry

Deep learning models are seeing increased use as methods to predict mutational effects or allowed mutations in proteins. The models commonly used for these purposes include large language models (LLMs) and 3D Convolutional Neural Networks (CNNs). These two model types have very different architectures and are commonly trained on different representations of proteins. LLMs make use of the transformer architecture and are trained purely on protein sequences whereas 3D CNNs are trained on voxelized representations of local protein structure. While comparable overall prediction accuracies have been reported for both types of models, it is not known to what extent these models make comparable specific predictions and/or generalize protein biochemistry in similar ways. Here, we perform a systematic comparison of two LLMs and two structure-based models (CNNs) and show that the different model types have distinct strengths and weaknesses. The overall prediction accuracies are largely uncorrelated between the sequence- and structure-based models. Overall, the two structure-based models are better at predicting buried aliphatic and hydrophobic residues whereas the two LLMs are better at predicting solvent-exposed polar and charged amino acids. Finally, we find that a combined model that takes the individual model predictions as input can leverage these individual model strengths and results in significantly improved overall prediction accuracy.

Two sequence- and two structure-based ML models have learned different aspects of protein biochemistry

Deep learning models are seeing increased use as methods to predict mutational effects or allowed mutations in proteins. The models commonly used for these purposes include large language models (LLMs) and 3D Convolutional Neural Networks (CNNs). These two model types have very different architectures and are commonly trained on different representations of proteins. LLMs make use of the transformer architecture and are trained purely on protein sequences whereas 3D CNNs are trained on voxelized representations of local protein structure. While comparable overall prediction accuracies have been reported for both types of models, it is not known to what extent these models make comparable specific predictions and/or generalize protein biochemistry in similar ways. Here, we perform a systematic comparison of two LLMs and two structure-based models (CNNs) and show that the different model types have distinct strengths and weaknesses. The overall prediction accuracies are largely uncorrelated between the sequence- and structure-based models. Overall, the two structure-based models are better at predicting buried aliphatic and hydrophobic residues whereas the two LLMs are better at predicting solvent-exposed polar and charged amino acids. Finally, we find that a combined model that takes the individual model predictions as input can leverage these individual model strengths and results in significantly improved overall prediction accuracy.

Understanding structure-guided variant effect predictions using 3D convolutional neural networks

Predicting pathogenicity of missense variants in molecular diagnostics remains a challenge despite the available wealth of data, such as evolutionary information, and the wealth of tools to integrate that data. We describe DeepRank-Mut, a configurable framework designed to extract and learn from physicochemically relevant features of amino acids surrounding missense variants in 3D space. For each variant, various atomic and residue-level features are extracted from its structural environment, including sequence conservation scores of the surrounding amino acids, and stored in multi-channel 3D voxel grids which are then used to train a 3D convolutional neural network (3D-CNN). The resultant model gives a probabilistic estimate of whether a given input variant is disease-causing or benign. We find that the performance of our 3D-CNN model, on independent test datasets, is comparable to other widely used resources which also combine sequence and structural features. Based on the 10-fold cross-validation experiments, we achieve an average accuracy of 0.77 on the independent test datasets. We discuss the contribution of the variant neighborhood in the model's predictive power, in addition to the impact of individual features on the model's performance. Two key features: evolutionary information of residues in the variant neighborhood and their solvent accessibilities were observed to influence the predictions. We also highlight how predictions are impacted by the underlying disease mechanisms of missense mutations and offer insights into understanding these to improve pathogenicity predictions. Our study presents aspects to take into consideration when adopting deep learning approaches for protein structure-guided pathogenicity predictions.

Metal3D: a general deep learning framework for accurate metal ion location prediction in proteins

Metal ions are essential cofactors for many proteins and play a crucial role in many applications such as enzyme design or design of protein-protein interactions because they are biologically abundant, tether to the protein using strong interactions, and have favorable catalytic properties. Computational design of metalloproteins is however hampered by the complex electronic structure of many biologically relevant metals such as zinc . In this work, we develop two tools - Metal3D (based on 3D convolutional neural networks) and Metal1D (solely based on geometric criteria) to improve the location prediction of zinc ions in protein structures. Comparison with other currently available tools shows that Metal3D is the most accurate zinc ion location predictor to date with predictions within 0.70 ± 0.64 Å of experimental locations. Metal3D outputs a confidence metric for each predicted site and works on proteins with few homologes in the protein data bank. Metal3D predicts a global zinc density that can be used for annotation of computationally predicted structures and a per residue zinc density that can be used in protein design workflows. Currently trained on zinc, the framework of Metal3D is readily extensible to other metals by modifying the training data.

OPUS-Mut: Studying the Effect of Protein Mutation through Side-Chain Modeling

Predicting the effect of protein mutation is crucial in many applications such as protein design, protein evolution, and genetic disease analysis. Structurally, mutation is basically the replacement of the side chain of a particular residue. Therefore, accurate side-chain modeling is useful in studying the effect of mutation. Here, we propose a computational method, namely, OPUS-Mut, which significantly outperforms other backbone-dependent side-chain modeling methods including our previous method OPUS-Rota4. We evaluate OPUS-Mut by four case studies on Myoglobin, p53, HIV-1 protease, and T4 lysozyme. The results show that the predicted structures of side chains of different mutants are consistent well with their experimentally determined results. In addition, when the residues with significant structural shifts upon the mutation are considered, it is found that the extent of the predicted structural shift of these affected residues can be correlated reasonably well with the functional changes of the mutant measured by experiments. OPUS-Mut can also help one to identify the harmful and benign mutations and thus may guide the construction of a protein with relatively low sequence homology but with a similar structure.