DeeplyTough: Learning Structural Comparison of Protein Binding Sites

Systematic Investigation of Dual-Target-Directed Ligands

Multitarget-directed ligands (MTDLs) are compounds rationally designed to affect multiple targets, aiming for a better therapeutic profile. For over 20 years, MTDLs have garnered increasing attention, the idea being that their full potential would have been achieved, thanks to unprecedented target combinations and advanced medicinal chemistry strategies. This study presents a literature mining effort resulting in a data set of dual-target-directed ligands (DTDLs), the fundamental example of MTDLs. We used this data set to evaluate the rationale behind target selection and the chemical novelty of DTDLs targeting specific protein combinations. Our analysis focused on DTDL targets in terms of biological function, disease association, structure, and chemogenomic traits. We also compared DTDLs with single-target compounds. We found that well-known target pathology associations often guide DTDL design, leveraging existing chemical scaffolds and binding pocket similarities. These findings highlight the current state of the field and suggest substantial untapped potential for rational polypharmacology.

Development of a novel representation of drug 3D structures and enhancement of the TSR-based method for probing drug and target interactions

Understanding the mechanisms underlying interactions between drugs and target proteins is critical for drug discovery. In our earlier studies, we introduced the Triangular Spatial Relationship (TSR)-based algorithm, which enables the representation of a protein's 3D structure as a vector of integers (TSR keys). These TSR keys correspond to substructures of the 3D structure of a protein and are computed based on the triangles constructed by all possible triples of Cα atoms within the protein. In this study, we report on a new TSR-based algorithm for probing drug and target interactions. Specifically, we have extended the previous algorithm in three novel directions: TSR keys for representing the 3D structure of a drug or a ligand, cross TSR keys between drugs and their targets and intra-residual TSR keys for phosphorylated amino acids. The outcomes illustrate the key contributions as follows: (i) The TSR-based method, which uses the TSR keys as features, is unique in its capability to interpret hierarchical relationships of drugs as well as drug - target complexes using common and specific TSR keys. (ii) The method can distinguish not only the binding sites from the rest of the protein structures, but also the binding sites of primary targets from those of off-targets. (iii) The method has the potential to correlate the 3D structures of drugs with their functions. (iv) Representation of 3D structures by TSR keys has its unique advantage in terms of ease of making searching for similar substructures across structure datasets easier. In summary, this study presents a novel computational methodology, with significant advantages, for providing insights into the mechanism underlying drug and target interactions.

LoCoHD: a metric for comparing local environments of proteins

Protein folds and the local environments they create can be compared using a variety of differently designed measures, such as the root mean squared deviation, the global distance test, the template modeling score or the local distance difference test. Although these measures have proven to be useful for a variety of tasks, each fails to fully incorporate the valuable chemical information inherent to atoms and residues, and considers these only partially and indirectly. Here, we develop the highly flexible local composition Hellinger distance (LoCoHD) metric, which is based on the chemical composition of local residue environments. Using LoCoHD, we analyze the chemical heterogeneity of amino acid environments and identify valines having the most conserved-, and arginines having the most variable chemical environments. We use LoCoHD to investigate structural ensembles, to evaluate critical assessment of structure prediction (CASP) competitors, to compare the results with the local distance difference test (lDDT) scoring system, and to evaluate a molecular dynamics simulation. We show that LoCoHD measurements provide unique information about protein structures that is distinct from, for example, those derived using the alignment-based RMSD metric, or the similarly distance matrix-based but alignment-free lDDT metric.

SiteMine: Large-scale binding site similarity searching in protein structure databases

Drug discovery and design challenges, such as drug repurposing, analyzing protein-ligand and protein-protein complexes, ligand promiscuity studies, or function prediction, can be addressed by protein binding site similarity analysis. Although numerous tools exist, they all have individual strengths and drawbacks with regard to run time, provision of structure superpositions, and applicability to diverse application domains. Here, we introduce SiteMine, an all-in-one database-driven, alignment-providing binding site similarity search tool to tackle the most pressing challenges of binding site comparison. The performance of SiteMine is evaluated on the ProSPECCTs benchmark, showing a promising performance on most of the data sets. The method performs convincingly regarding all quality criteria for reliable binding site comparison, offering a novel state-of-the-art approach for structure-based molecular design based on binding site comparisons. In a SiteMine showcase, we discuss the high structural similarity between cathepsin L and calpain 1 binding sites and give an outlook on the impact of this finding on structure-based drug design. SiteMine is available at https://uhh.de/naomi.

Graph Attention Site Prediction (GrASP): Identifying Druggable Binding Sites Using Graph Neural Networks with Attention

Identifying and discovering druggable protein binding sites is an important early step in computer-aided drug discovery, but it remains a difficult task where most campaigns rely on a priori knowledge of binding sites from experiments. Here, we present a binding site prediction method called Graph Attention Site Prediction (GrASP) and re-evaluate assumptions in nearly every step in the site prediction workflow from data set preparation to model evaluation. GrASP is able to achieve state-of-the-art performance at recovering binding sites in PDB structures while maintaining a high degree of precision which will minimize wasted computation in downstream tasks such as docking and free energy perturbation.

VirtuousPocketome: a computational tool for screening protein-ligand complexes to identify similar binding sites

Protein residues within binding pockets play a critical role in determining the range of ligands that can interact with a protein, influencing its structure and function. Identifying structural similarities in proteins offers valuable insights into their function and activation mechanisms, aiding in predicting protein-ligand interactions, anticipating off-target effects, and facilitating the development of therapeutic agents. Numerous computational methods assessing global or local similarity in protein cavities have emerged, but their utilization is impeded by complexity, impractical automation for amino acid pattern searches, and an inability to evaluate the dynamics of scrutinized protein-ligand systems. Here, we present a general, automatic and unbiased computational pipeline, named VirtuousPocketome, aimed at screening huge databases of proteins for similar binding pockets starting from an interested protein-ligand complex. We demonstrate the pipeline's potential by exploring a recently-solved human bitter taste receptor, i.e. the TAS2R46, complexed with strychnine. We pinpointed 145 proteins sharing similar binding sites compared to the analysed bitter taste receptor and the enrichment analysis highlighted the related biological processes, molecular functions and cellular components. This work represents the foundation for future studies aimed at understanding the effective role of tastants outside the gustatory system: this could pave the way towards the rationalization of the diet as a supplement to standard pharmacological treatments and the design of novel tastants-inspired compounds to target other proteins involved in specific diseases or disorders. The proposed pipeline is publicly accessible, can be applied to any protein-ligand complex, and could be expanded to screen any database of protein structures.

Deep learning for protein structure prediction and design-progress and applications

Proteins are the key molecular machines that orchestrate all biological processes of the cell. Most proteins fold into three-dimensional shapes that are critical for their function. Studying the 3D shape of proteins can inform us of the mechanisms that underlie biological processes in living cells and can have practical applications in the study of disease mutations or the discovery of novel drug treatments. Here, we review the progress made in sequence-based prediction of protein structures with a focus on applications that go beyond the prediction of single monomer structures. This includes the application of deep learning methods for the prediction of structures of protein complexes, different conformations, the evolution of protein structures and the application of these methods to protein design. These developments create new opportunities for research that will have impact across many areas of biomedical research.

Drugging the entire human proteome: Are we there yet?

Each of the ∼20,000 proteins in the human proteome is a potential target for compounds that bind to it and modify its function. The 3D structures of most of these proteins are now available. Here, we discuss the prospects for using these structures to perform proteome-wide virtual HTS (VHTS). We compare physics-based (docking) and AI VHTS approaches, some of which are now being applied with large databases of compounds to thousands of targets. Although preliminary proteome-wide screens are now within our grasp, further methodological developments are expected to improve the accuracy of the results.

Prop3D: A flexible, Python-based platform for machine learning with protein structural properties and biophysical data

BACKGROUND

Machine learning (ML) has a rich history in structural bioinformatics, and modern approaches, such as deep learning, are revolutionizing our knowledge of the subtle relationships between biomolecular sequence, structure, function, dynamics and evolution. As with any advance that rests upon statistical learning approaches, the recent progress in biomolecular sciences is enabled by the availability of vast volumes of sufficiently-variable data. To be useful, such data must be well-structured, machine-readable, intelligible and manipulable. These and related requirements pose challenges that become especially acute at the computational scales typical in ML. Furthermore, in structural bioinformatics such data generally relate to protein three-dimensional (3D) structures, which are inherently more complex than sequence-based data. A significant and recurring challenge concerns the creation of large, high-quality, openly-accessible datasets that can be used for specific training and benchmarking tasks in ML pipelines for predictive modeling projects, along with reproducible splits for training and testing.

RESULTS

Here, we report 'Prop3D', a platform that allows for the creation, sharing and extensible reuse of libraries of protein domains, featurized with biophysical and evolutionary properties that can range from detailed, atomically-resolved physicochemical quantities (e.g., electrostatics) to coarser, residue-level features (e.g., phylogenetic conservation). As a community resource, we also supply a 'Prop3D-20sf' protein dataset, obtained by applying our approach to CATH . We have developed and deployed the Prop3D framework, both in the cloud and on local HPC resources, to systematically and reproducibly create comprehensive datasets via the Highly Scalable Data Service ( HSDS ). Our datasets are freely accessible via a public HSDS instance, or they can be used with accompanying Python wrappers for popular ML frameworks.

CONCLUSION

Prop3D and its associated Prop3D-20sf dataset can be of broad utility in at least three ways. Firstly, the Prop3D workflow code can be customized and deployed on various cloud-based compute platforms, with scalability achieved largely by saving the results to distributed HDF5 files via HSDS . Secondly, the linked Prop3D-20sf dataset provides a hand-crafted, already-featurized dataset of protein domains for 20 highly-populated CATH families; importantly, provision of this pre-computed resource can aid the more efficient development (and reproducible deployment) of ML pipelines. Thirdly, Prop3D-20sf's construction explicitly takes into account (in creating datasets and data-splits) the enigma of 'data leakage', stemming from the evolutionary relationships between proteins.

DeepGRID: Deep Learning Using GRID Descriptors for BBB Prediction

Deep Learning approaches are able to automatically extract relevant features from the input data and capture nonlinear relationships between the input and output. In this work, we present the GRID-derived AI (GrAId) descriptors, a simple modification to GRID MIFs that facilitate their use in combination with Convolutional Neural Networks (CNNs) to build Deep Learning models in a rotationally, conformationally, and alignment-independent approach we are calling DeepGRID. To our knowledge, this is the first time that GRID MIFs have been combined with CNNs in a Deep Learning approach. We applied the approach to build regression and classification models for blood-brain barrier permeation, an important factor when designing CNS drugs and conversely when designing to avoid off-target effects for CNS-inactive drugs. The VolSurf approach was one of the first to successfully model this property from three-dimensional structures, using descriptors derived from their GRID Molecular Interaction Fields (MIFs) in combination with PLS. We compared the DeepGRID models with others built using the hand-crafted VolSurf descriptors in combination with both PLS and Random Forest (RF). Both the DeepGRID and RF regression models performed best according to the % of compounds with a Geometric Mean Fold Error (GMFE) within 2-fold of the experimental data. Applying these regression models as classifiers, for the smaller 332 and 416 compound data sets all models performed well with ROC AUC values of ∼0.9 on the external test set. For the larger 2105 compound data set, the DeepGRID classifier performed the best with an AUC of 0.87 on the external test set with the RF model having an AUC of 0.84 and the original VolSurf lgBB model having an AUC of 0.83.

PocketAnchor: Learning structure-based pocket representations for protein-ligand interaction prediction

Protein-ligand interactions are essential for cellular activities and drug discovery processes. Appropriately and effectively representing protein features is of vital importance for developing computational approaches, especially data-driven methods, for predicting protein-ligand interactions. However, existing approaches may not fully investigate the features of the ligand-occupying regions in the protein pockets. Here, we design a structure-based protein representation method, named PocketAnchor, for capturing the local environmental and spatial features of protein pockets to facilitate protein-ligand interaction-related learning tasks. We define "anchors" as probe points reaching into the cavities and those located near the surface of proteins, and we design a specific message passing strategy for gathering local information from the atoms and surface neighboring these anchors. Comprehensive evaluation of our method demonstrated its successful applications in pocket detection and binding affinity prediction, which indicated that our anchor-based approach can provide effective protein feature representations for improving the prediction of protein-ligand interactions.

Graph Attention Site Prediction (GrASP): Identifying Druggable Binding Sites Using Graph Neural Networks with Attention

Identifying and discovering druggable protein binding sites is an important early step in computer-aided drug discovery but remains a difficult task where most campaigns rely on a priori knowledge of binding sites from experiments. Here we present a novel binding site prediction method called Graph Attention Site Prediction (GrASP) and re-evaluate assumptions in nearly every step in the site prediction workflow from dataset preparation to model evaluation. GrASP is able to achieve state-of-the-art performance at recovering binding sites in PDB structures while maintaining a high degree of precision which will minimize wasted computation in downstream tasks such as docking and free energy perturbation.

HLA Class I Supertype Classification Based on Structural Similarity

HLA class I proteins, a critical component in adaptive immunity, bind and present intracellular Ags to CD8+ T cells. The extreme polymorphism of HLA genes and associated peptide binding specificities leads to challenges in various endeavors, including neoantigen vaccine development, disease association studies, and HLA typing. Supertype classification, defined by clustering functionally similar HLA alleles, has proven helpful in reducing the complexity of distinguishing alleles. However, determining supertypes via experiments is impractical, and current in silico classification methods exhibit limitations in stability and functional relevance. In this study, by incorporating three-dimensional structures we present a method for classifying HLA class I molecules with improved breadth, accuracy, stability, and flexibility. Critical for these advances is our finding that structural similarity highly correlates with peptide binding specificity. The new classification should be broadly useful in peptide-based vaccine development and HLA-disease association studies.

Classification of Protein-Binding Sites Using a Spherical Convolutional Neural Network

The analysis and comparison of protein-binding sites aid various applications in the drug discovery process, e.g., hit finding, drug repurposing, and polypharmacology. Classification of binding sites has been a hot topic for the past 30 years, and many different methods have been published. The rapid development of machine learning computational algorithms, coupled with the large volume of publicly available protein-ligand 3D structures, makes it possible to apply deep learning techniques in binding site comparison. Our method uses a cutting-edge spherical convolutional neural network based on the DeepSphere architecture to learn global representations of protein-binding sites. The model was trained on TOUGH-C1 and TOUGH-M1 data and validated with the ProSPECCTs datasets. Our results show that our model can (1) perform well in protein-binding site similarity and classification tasks and (2) learn and separate the physicochemical properties of binding sites. Lastly, we tested the model on a set of kinases, where the results show that it is able to cluster the different kinase subfamilies effectively. This example demonstrates the method's promise for lead hopping within or outside a protein target, directly based on binding site information.

Holo Protein Conformation Generation from Apo Structures by Ligand Binding Site Refinement

An important part in structure-based drug design is the selection of an appropriate protein structure. It has been revealed that a holo protein structure that contains a well-defined binding site is a much better choice than an apo structure in structure-based drug discovery. Therefore, it is valuable to obtain a holo-like protein conformation from apo structures in the case where no holo structure is available. Meeting the need, we present a robust approach to generate reliable holo-like structures from apo structures by ligand binding site refinement with restraints derived from holo templates with low homology. Our method was tested on a test set of 32 proteins from the DUD-E data set and compared with other approaches. It was shown that our method successfully refined the apo structures toward the corresponding holo conformations for 23 of 32 proteins, reducing the average all-heavy-atom RMSD of binding site residues by 0.48 Å. In addition, when evaluated against all the holo structures in the protein data bank, our method can improve the binding site RMSD for 14 of 19 cases that experience significant conformational changes. Furthermore, our refined structures also demonstrate their advantages over the apo structures in ligand binding mode predictions by both rigid docking and flexible docking and in virtual screening on the database of active and decoy ligands from the DUD-E. These results indicate that our method is effective in recovering holo-like conformations and will be valuable in structure-based drug discovery.

A multilayer dynamic perturbation analysis method for predicting ligand-protein interactions

BACKGROUND

Ligand-protein interactions play a key role in defining protein function, and detecting natural ligands for a given protein is thus a very important bioengineering task. In particular, with the rapid development of AI-based structure prediction algorithms, batch structural models with high reliability and accuracy can be obtained at low cost, giving rise to the urgent requirement for the prediction of natural ligands based on protein structures. In recent years, although several structure-based methods have been developed to predict ligand-binding pockets and ligand-binding sites, accurate and rapid methods are still lacking, especially for the prediction of ligand-binding regions and the spatial extension of ligands in the pockets.

RESULTS

In this paper, we proposed a multilayer dynamics perturbation analysis (MDPA) method for predicting ligand-binding regions based solely on protein structure, which is an extended version of our previously developed fast dynamic perturbation analysis (FDPA) method. In MDPA/FDPA, ligand binding tends to occur in regions that cause large changes in protein conformational dynamics. MDPA, examined using a standard validation dataset of ligand-protein complexes, yielded an averaged ligand-binding site prediction Matthews coefficient of 0.40, with a prediction precision of at least 50% for 71% of the cases. In particular, for 80% of the cases, the predicted ligand-binding region overlaps the natural ligand by at least 50%. The method was also compared with other state-of-the-art structure-based methods.

CONCLUSIONS

MDPA is a structure-based method to detect ligand-binding regions on protein surface. Our calculations suggested that a range of spaces inside the protein pockets has subtle interactions with the protein, which can significantly impact on the overall dynamics of the protein. This work provides a valuable tool as a starting point upon which further docking and analysis methods can be used for natural ligand detection in protein functional annotation. The source code of MDPA method is freely available at: https://github.com/mingdengming/mdpa .

Estimating the Similarity between Protein Pockets

With the exponential increase in publicly available protein structures, the comparison of protein binding sites naturally emerged as a scientific topic to explain observations or generate hypotheses for ligand design, notably to predict ligand selectivity for on- and off-targets, explain polypharmacology, and design target-focused libraries. The current review summarizes the state-of-the-art computational methods applied to pocket detection and comparison as well as structural druggability estimates. The major strengths and weaknesses of current pocket descriptors, alignment methods, and similarity search algorithms are presented. Lastly, an exhaustive survey of both retrospective and prospective applications in diverse medicinal chemistry scenarios illustrates the capability of the existing methods and the hurdle that still needs to be overcome for more accurate predictions.

GraphSite: Ligand Binding Site Classification with Deep Graph Learning

The binding of small organic molecules to protein targets is fundamental to a wide array of cellular functions. It is also routinely exploited to develop new therapeutic strategies against a variety of diseases. On that account, the ability to effectively detect and classify ligand binding sites in proteins is of paramount importance to modern structure-based drug discovery. These complex and non-trivial tasks require sophisticated algorithms from the field of artificial intelligence to achieve a high prediction accuracy. In this communication, we describe GraphSite, a deep learning-based method utilizing a graph representation of local protein structures and a state-of-the-art graph neural network to classify ligand binding sites. Using neural weighted message passing layers to effectively capture the structural, physicochemical, and evolutionary characteristics of binding pockets mitigates model overfitting and improves the classification accuracy. Indeed, comprehensive cross-validation benchmarks against a large dataset of binding pockets belonging to 14 diverse functional classes demonstrate that GraphSite yields the class-weighted F1-score of 81.7%, outperforming other approaches such as molecular docking and binding site matching. Further, it also generalizes well to unseen data with the F1-score of 70.7%, which is the expected performance in real-world applications. We also discuss new directions to improve and extend GraphSite in the future.

Exploring kinase family inhibitors and their moiety preferences using deep SHapley additive exPlanations

Background

While it has been known that human protein kinases mediate most signal transductions in cells and their dysfunction can result in inflammatory diseases and cancers, it remains a challenge to find effective kinase inhibitor as drugs for these diseases. One major challenge is the compensatory upregulation of related kinases following some critical kinase inhibition. To circumvent the compensatory effect, it is desirable to have inhibitors that inhibit all the kinases belonging to the same family, instead of targeting only a few kinases. However, finding inhibitors that target a whole kinase family is laborious and time consuming in wet lab.

Results

In this paper, we present a computational approach taking advantage of interpretable deep learning models to address this challenge. Specifically, we firstly collected 9,037 inhibitor bioassay results (with 3991 active and 5046 inactive pairs) for eight kinase families (including EGFR, Jak, GSK, CLK, PIM, PKD, Akt and PKG) from the ChEMBL25 Database and the Metz Kinase Profiling Data. We generated 238 binary moiety features for each inhibitor, and used the features as input to train eight deep neural networks (DNN) models to predict whether an inhibitor is active for each kinase family. We then employed the SHapley Additive exPlanations (SHAP) to analyze the importance of each moiety feature in each classification model, identifying moieties that are in the common kinase hinge sites across the eight kinase families, as well as moieties that are specific to some kinase families. We finally validated these identified moieties using experimental crystal structures to reveal their functional importance in kinase inhibition.

Conclusion

With the SHAP methodology, we identified two common moieties for eight kinase families, 9 EGFR-specific moieties, and 6 Akt-specific moieties, that bear functional importance in kinase inhibition. Our result suggests that SHAP has the potential to help finding effective pan-kinase family inhibitors.

Improving ΔΔG Predictions with a Multitask Convolutional Siamese Network

The lead optimization phase of drug discovery refines an initial hit molecule for desired properties, especially potency. Synthesis and experimental testing of the small perturbations during this refinement can be quite costly and time-consuming. Relative binding free energy (RBFE, also referred to as ΔΔG) methods allow the estimation of binding free energy changes after small changes to a ligand scaffold. Here, we propose and evaluate a Siamese convolutional neural network (CNN) for the prediction of RBFE between two bound ligands. We show that our multitask loss is able to improve on a previous state-of-the-art Siamese network for RBFE prediction via increased regularization of the latent space. The Siamese network architecture is well suited to the prediction of RBFE in comparison to a standard CNN trained on the same data (Pearson's R of 0.553 and 0.5, respectively). When evaluated on a left-out protein family, our Siamese CNN shows variability in its RBFE predictive performance depending on the protein family being evaluated (Pearson's R ranging from -0.44 to 0.97). RBFE prediction performance can be improved during generalization by injecting only a few examples (few-shot learning) from the evaluation data set during model training.

Structure-based drug repurposing: traditional and advanced AI/ML-aided methods

The current global health emergency in the form of the Coronavirus 2019 (COVID-19) pandemic has highlighted the need for fast, accurate, and efficient drug discovery pipelines. Traditional drug discovery projects relying on in vitro high-throughput screening (HTS) involve large investments and sophisticated experimental set-ups, affordable only to big biopharmaceutical companies. In this scenario, application of efficient state-of-the-art computational methods and modern artificial intelligence (AI)-based algorithms for rapid screening of repurposable chemical space [approved drugs and natural products (NPs) with proven pharmacokinetic profiles] to identify the initial leads is a powerful option to save resources and time. Structure-based drug repurposing is a popular in silico repurposing approach. In this review, we discuss traditional and modern AI-based computational methods and tools applied at various stages for structure-based drug discovery (SBDD) pipelines. Additionally, we highlight the role of generative models in generating molecules with scaffolds from repurposable chemical space. Teaser: This review highlights the importance of repurposable chemical space, and the contributions of conventional in silico approaches and modern machine-learning algorithms for rapid structure-based drug repurposing.

Pocket2Drug: An Encoder-Decoder Deep Neural Network for the Target-Based Drug Design

Computational modeling is an essential component of modern drug discovery. One of its most important applications is to select promising drug candidates for pharmacologically relevant target proteins. Because of continuing advances in structural biology, putative binding sites for small organic molecules are being discovered in numerous proteins linked to various diseases. These valuable data offer new opportunities to build efficient computational models predicting binding molecules for target sites through the application of data mining and machine learning. In particular, deep neural networks are powerful techniques capable of learning from complex data in order to make informed drug binding predictions. In this communication, we describe Pocket2Drug, a deep graph neural network model to predict binding molecules for a given a ligand binding site. This approach first learns the conditional probability distribution of small molecules from a large dataset of pocket structures with supervised training, followed by the sampling of drug candidates from the trained model. Comprehensive benchmarking simulations show that using Pocket2Drug significantly improves the chances of finding molecules binding to target pockets compared to traditional drug selection procedures. Specifically, known binders are generated for as many as 80.5% of targets present in the testing set consisting of dissimilar data from that used to train the deep graph neural network model. Overall, Pocket2Drug is a promising computational approach to inform the discovery of novel biopharmaceuticals.

Three-Dimensional Convolutional Neural Networks Utilizing Molecular Topological Features for Accurate Atomization Energy Predictions

Deep learning methods provide a novel way to establish a correlation between two quantities. In this context, computer vision techniques such as three-dimensional (3D)-convolutional neural networks become a natural choice to associate a molecular property with its structure due to the inherent 3D nature of a molecule. However, traditional 3D input data structures are intrinsically sparse in nature, which tend to induce instabilities during the learning process, which in turn may lead to underfitted results. To address this deficiency, in this project, we propose to use quantum-chemically derived molecular topological features, namely, localized orbital locator and electron localization function, as molecular descriptors, which provide a relatively denser input representation in a 3D space. Such topological features provide a detailed picture of the atomic and electronic configuration and interatomic interactions in the molecule and hence are ideal for predicting properties that are highly dependent on the physical or electronic structure of the molecule. Herein, we demonstrate the efficacy of our proposed model by applying it to the task of predicting atomization energies for the QM9-G4MP2 data set, which contains ∼134k molecules. Furthermore, we incorporated the Δ-machine learning approach into our model, which enabled us to reach beyond benchmark accuracy levels (∼1.0 kJ mol^-1). As a result, we consistently obtain impressive mean absolute errors of the order 0.1 kcal mol^-1 (∼0.42 kJ mol^-1) versus the G4(MP2) theory using relatively modest models, which could potentially be improved further in a systematic manner using additional compute resources.

The Signaling Pathway That cGAMP Riboswitches Found: Analysis and Application of Riboswitches to Study cGAMP Signaling in Geobacter sulfurreducens

The Hypr cGAMP signaling pathway was discovered via the function of the riboswitch. In this study, we show the development of a method for affinity capture followed by sequencing to identify non-coding RNA regions that bind nucleotide signals such as cGAMP. The RNAseq of affinity-captured cGAMP riboswitches from the Geobacter sulfurreducens transcriptome highlights general challenges that remain for this technique. Furthermore, by applying riboswitch reporters in vivo, we identify new growth conditions and transposon mutations that affect cGAMP levels in G. sulfurreducens. This work reveals an extensive regulatory network and supports a second functional cGAMP synthase gene in G. sulfurreducens. The activity of the second synthase was validated using riboswitch-based fluorescent biosensors, and is the first known example of an active enzyme with a variant GGDDF motif.

Enhancing preclinical drug discovery with artificial intelligence

Artificial intelligence (AI) is becoming an integral part of drug discovery. It has the potential to deliver across the drug discovery and development value chain, starting from target identification and reaching through clinical development. In this review, we provide an overview of current AI technologies and a glimpse of how AI is reimagining preclinical drug discovery by highlighting examples where AI has made a real impact. Considering the excitement and hyperbole surrounding AI in drug discovery, we aim to present a realistic view by discussing both opportunities and challenges in adopting AI in drug discovery.

Artificial intelligence for the discovery of novel antimicrobial agents for emerging infectious diseases

The search for effective drugs to treat new and existing diseases is a laborious one requiring a large investment of capital, resources, and time. The coronavirus 2019 (COVID-19) pandemic has been a painful reminder of the lack of development of new antimicrobial agents to treat emerging infectious diseases. Artificial intelligence (AI) and other in silico techniques can drive a more efficient, cost-friendly approach to drug discovery by helping move potential candidates with better clinical tolerance forward in the pipeline. Several research teams have developed successful AI platforms for hit identification, lead generation, and lead optimization. In this review, we investigate the technologies at the forefront of spearheading an AI revolution in drug discovery and pharmaceutical sciences.

Connecting chemistry and biology through molecular descriptors

Through the representation of small molecule structures as numerical descriptors and the exploitation of the similarity principle, chemoinformatics has made paramount contributions to drug discovery, from unveiling mechanisms of action and repurposing approved drugs to de novo crafting of molecules with desired properties and tailored targets. Yet, the inherent complexity of biological systems has fostered the implementation of large-scale experimental screenings seeking a deeper understanding of the targeted proteins, the disrupted biological processes and the systemic responses of cells to chemical perturbations. After this wealth of data, a new generation of data-driven descriptors has arisen providing a rich portrait of small molecule characteristics that goes beyond chemical properties. Here, we give an overview of biologically relevant descriptors, covering chemical compounds, proteins and other biological entities, such as diseases and cell lines, while aligning them to the major contributions in the field from disciplines, such as natural language processing or computer vision. We now envision a new scenario for chemical and biological entities where they both are translated into a common numerical format. In this computational framework, complex connections between entities can be unveiled by means of simple arithmetic operations, such as distance measures, additions, and subtractions.

DeepPocket: Ligand Binding Site Detection and Segmentation using 3D Convolutional Neural Networks

A structure-based drug design pipeline involves the development of potential drug molecules or ligands that form stable complexes with a given receptor at its binding site. A prerequisite to this is finding druggable and functionally relevant binding sites on the 3D structure of the protein. Although several methods for detecting binding sites have been developed beforehand, a majority of them surprisingly fail in the identification and ranking of binding sites accurately. The rapid adoption and success of deep learning algorithms in various sections of structural biology beckons the usage of such algorithms for accurate binding site detection. As a combination of geometry based software and deep learning, we report a novel framework, DeepPocket that utilizes 3D convolutional neural networks for the rescoring of pockets identified by Fpocket and further segments these identified cavities on the protein surface. Apart from this, we also propose another data set SC6K containing protein structures submitted in the Protein Data Bank (PDB) from January 1st, 2018, until February 28th, 2020, for ligand binding site (LBS) detection. DeepPocket's results on various binding site data sets and SC6K highlight its better performance over current state-of-the-art methods and good generalization ability over novel structures.

Machine learning and big data provide crucial insight for future biomaterials discovery and research

Machine learning have been widely adopted in a variety of fields including engineering, science, and medicine revolutionizing how data is collected, used, and stored. Their implementation has led to a drastic increase in the number of computational models for the prediction of various numerical, categorical, or association events given input variables. We aim to examine recent advances in the use of machine learning when applied to the biomaterial field. Specifically, quantitative structure properties relationships offer the unique ability to correlate microscale molecular descriptors to larger macroscale material properties. These new models can be broken down further into four categories: regression, classification, association, and clustering. We examine recent approaches and new uses of machine learning in the three major categories of biomaterials: metals, polymers, and ceramics for rapid property prediction and trend identification. While current research is promising, limitations in the form of lack of standardized reporting and available databases complicates the implementation of described models. Herein, we hope to provide a snapshot of the current state of the field and a beginner's guide to navigating the intersection of biomaterials research and machine learning. STATEMENT OF SIGNIFICANCE: Machine learning and its methods have found a variety of uses beyond the field of computer science but have largely been neglected by those in realm of biomaterials. Through the use of more computational methods, biomaterials development can be expediated while reducing the need for standard trial and error methods. Within, we introduce four basic models that readers can potentially apply to their current research as well as current applications within the field. Furthermore, we hope that this article may act as a "call to action" for readers to realize and address the current lack of implementation within the biomaterials field.

A fast protein binding site comparison algorithm for proteome-wide protein function prediction and drug repurposing

The expansion of three-dimensional protein structures and enhanced computing power have significantly facilitated our understanding of protein sequence/structure/function relationships. A challenge in structural genomics is to predict the function of uncharacterized proteins. Protein function deconvolution based on global sequence or structural homology is impracticable when a protein relates to no other proteins with known function, and in such cases, functional relationships can be established by detecting their local ligand binding site similarity. Here, we introduce a sequence order-independent comparison algorithm, PocketShape, for structural proteome-wide exploration of protein functional site by fully considering the geometry of the backbones, orientation of the sidechains, and physiochemical properties of the pocket-lining residues. PocketShape is efficient in distinguishing similar from dissimilar ligand binding site pairs by retrieving 99.3% of the similar pairs while rejecting 100% of the dissimilar pairs on a dataset containing 1538 binding site pairs. This method successfully classifies 83 enzyme structures with diverse functions into 12 clusters, which is highly in accordance with the actual structural classification of proteins classification. PocketShape also achieves superior performances than other methods in protein profiling based on experimental data. Potential new applications for representative SARS-CoV-2 drugs Remdesivir and 11a are predicted. The high accuracy and time-efficient characteristics of PocketShape will undoubtedly make it a promising complementary tool for proteome-wide protein function inference and drug repurposing study.

De Novo Molecule Design Through the Molecular Generative Model Conditioned by 3D Information of Protein Binding Sites

De novo molecule design through the molecular generative model has gained increasing attention in recent years. Here, a novel generative model was proposed by integrating the three-dimensional (3D) structural information of the protein binding pocket into the conditional RNN (cRNN) model to control the generation of drug-like molecules. In this model, the composition of the protein binding pocket is effectively characterized through a coarse-grain strategy and the 3D information of the pocket can be represented by the sorted eigenvalues of the Coulomb matrix (EGCM) of the coarse-grained atoms composing the binding pocket. In current work, we used our EGCM method and a previously reported binding pocket descriptor, DeeplyTough, to train cRNN models and evaluated their performance. It has been shown that the model trained with the constraint of protein environment information has a clear tendency on generating compounds with higher similarity to the original X-ray-bound ligand than the normal RNN model and also better docking scores. Our results demonstrate the potential application of the controlled generative model for the targeted molecule generation and guided exploration on the drug-like chemical space.

Recent advancements in enzyme-mediated crosslinkable hydrogels: In vivo-mimicking strategies

Enzymes play a central role in fundamental biological processes and have been traditionally used to trigger various processes. In recent years, enzymes have been used to tune biomaterial responses and modify the chemical structures at desired sites. These chemical modifications have allowed the fabrication of various hydrogels for tissue engineering and therapeutic applications. This review provides a comprehensive overview of recent advancements in the use of enzymes for hydrogel fabrication. Strategies to enhance the enzyme function and improve biocompatibility are described. In addition, we describe future opportunities and challenges for the production of enzyme-mediated crosslinkable hydrogels.

De novo molecular design and generative models

Molecular design strategies are integral to therapeutic progress in drug discovery. Computational approaches for de novo molecular design have been developed over the past three decades and, recently, thanks in part to advances in machine learning (ML) and artificial intelligence (AI), the drug discovery field has gained practical experience. Here, we review these learnings and present de novo approaches according to the coarseness of their molecular representation: that is, whether molecular design is modeled on an atom-based, fragment-based, or reaction-based paradigm. Furthermore, we emphasize the value of strong benchmarks, describe the main challenges to using these methods in practice, and provide a viewpoint on further opportunities for exploration and challenges to be tackled in the upcoming years.

Current directions in combining simulation-based macromolecular modeling approaches with deep learning

Introduction: Structure-guided drug discovery relies on accurate computational methods for modeling macromolecules. Simulations provide means of predicting macromolecular folds, of discovering function from structure, and of designing macromolecules to serve as drugs. Success rates are limited for any of these tasks, however. Recently, deep neural network-based methods have greatly enhanced the accuracy of predictions of protein structure from sequence, generating excitement about the potential impact of deep learning.Areas covered: This review introduces biologists to deep neural network architecture, surveys recent successes of deep learning in structure prediction, and discusses emerging deep learning-based approaches for structure-function analysis and design. Particular focus is given to the interplay between simulation-based and neural network-based approaches.Expert opinion: As deep learning grows integral to macromolecular modeling, simulation- and neural network-based approaches must grow more tightly interconnected. Modular software architecture must emerge allowing both types of tools to be combined with maximal versatility. Open sharing of code under permissive licenses will be essential. Although experiments will remain the gold standard for reliable information to guide drug discovery, we may soon see successful drug development projects based on high-accuracy predictions from algorithms that combine simulation with deep learning - the ultimate validation of this combination's power.

Deep Learning in Protein Structural Modeling and Design

Deep learning is catalyzing a scientific revolution fueled by big data, accessible toolkits, and powerful computational resources, impacting many fields, including protein structural modeling. Protein structural modeling, such as predicting structure from amino acid sequence and evolutionary information, designing proteins toward desirable functionality, or predicting properties or behavior of a protein, is critical to understand and engineer biological systems at the molecular level. In this review, we summarize the recent advances in applying deep learning techniques to tackle problems in protein structural modeling and design. We dissect the emerging approaches using deep learning techniques for protein structural modeling and discuss advances and challenges that must be addressed. We argue for the central importance of structure, following the "sequence → structure → function" paradigm. This review is directed to help both computational biologists to gain familiarity with the deep learning methods applied in protein modeling, and computer scientists to gain perspective on the biologically meaningful problems that may benefit from deep learning techniques.