3DGT-DDI: 3D graph and text based neural network for drug-drug interaction prediction.

MOTIVATION: Drug-drug interactions (DDIs) occur during the combination of drugs. Identifying potential DDI helps us to study the mechanism behind the combination medication or adverse reactions so as to avoid the side effects. Although many artificial intelligence methods predict and mine potential DDI, they ignore the 3D structure information of drug molecules and do not fully consider the contribution of molecular substructure in DDI. RESULTS: We proposed a new deep learning architecture, 3DGT-DDI, a model composed of a 3D graph neural network and pre-trained text attention mechanism. We used 3D molecular graph structure and position information to enhance the prediction ability of the model for DDI, which enabled us to deeply explore the effect of drug substructure on DDI relationship. The results showed that 3DGT-DDI outperforms other state-of-the-art baselines. It achieved an 84.48% macro F1 score in the DDIExtraction 2013 shared task dataset. Also, our 3D graph model proves its performance and explainability through weight visualization on the DrugBank dataset. 3DGT-DDI can help us better understand and identify potential DDI, thereby helping to avoid the side effects of drug mixing. AVAILABILITY: The source code and data are available at https://github.com/hehh77/3DGT-DDI.

FusionDTA: attention-based feature polymerizer and knowledge distillation for drug-target binding affinity prediction.

The prediction of drug-target affinity (DTA) plays an increasingly important role in drug discovery. Nowadays, lots of prediction methods focus on feature encoding of drugs and proteins, but ignore the importance of feature aggregation. However, the increasingly complex encoder networks lead to the loss of implicit information and excessive model size. To this end, we propose a deep-learning-based approach namely FusionDTA. For the loss of implicit information, a novel muti-head linear attention mechanism was utilized to replace the rough pooling method. This allows FusionDTA aggregates global information based on attention weights, instead of selecting the largest one as max-pooling does. To solve the redundancy issue of parameters, we applied knowledge distillation in FusionDTA by transfering learnable information from teacher model to student. Results show that FusionDTA performs better than existing models for the test domain on all evaluation metrics. We obtained concordance index (CI) index of 0.913 and 0.906 in Davis and KIBA dataset respectively, compared with 0.893 and 0.891 of previous state-of-art model. Under the cold-start constrain, our model proved to be more robust and more effective with unseen inputs than baseline methods. In addition, the knowledge distillation did save half of the parameters of the model, with only 0.006 reduction in CI index. Even FusionDTA with half the parameters could easily exceed the baseline on all metrics. In general, our model has superior performance and improves the effect of drug-target interaction (DTI) prediction. The visualization of DTI can effectively help predict the binding region of proteins during structure-based drug design.

NHGNN-DTA: a node-adaptive hybrid graph neural network for interpretable drug-target binding affinity prediction.

MOTIVATION: Large-scale prediction of drug-target affinity (DTA) plays an important role in drug discovery. In recent years, machine learning algorithms have made great progress in DTA prediction by utilizing sequence or structural information of both drugs and proteins. However, sequence-based algorithms ignore the structural information of molecules and proteins, while graph-based algorithms are insufficient in feature extraction and information interaction. RESULTS: In this article, we propose NHGNN-DTA, a node-adaptive hybrid neural network for interpretable DTA prediction. It can adaptively acquire feature representations of drugs and proteins and allow information to interact at the graph level, effectively combining the advantages of both sequence-based and graph-based approaches. Experimental results have shown that NHGNN-DTA achieved new state-of-the-art performance. It achieved the mean squared error (MSE) of 0.196 on the Davis dataset (below 0.2 for the first time) and 0.124 on the KIBA dataset (3% improvement). Meanwhile, in the case of cold start scenario, NHGNN-DTA proved to be more robust and more effective with unseen inputs than baseline methods. Furthermore, the multi-head self-attention mechanism endows the model with interpretability, providing new exploratory insights for drug discovery. The case study on Omicron variants of SARS-CoV-2 illustrates the efficient utilization of drug repurposing in COVID-19. AVAILABILITY AND IMPLEMENTATION: The source code and data are available at https://github.com/hehh77/NHGNN-DTA.

MRI-based automated machine learning model for preoperative identification of variant histology in muscle-invasive bladder carcinoma.

OBJECTIVES: It is essential yet highly challenging to preoperatively diagnose variant histologies such as urothelial carcinoma with squamous differentiation (UC w/SD) from pure UC in patients with muscle-invasive bladder carcinoma (MIBC), as their treatment strategy varies significantly. We developed a non-invasive automated machine learning (AutoML) model to preoperatively differentiate UC w/SD from pure UC in patients with MIBC. METHODS: A total of 119 MIBC patients who underwent baseline bladder MRI were enrolled in this study, including 38 patients with UC w/SD and 81 patients with pure UC. These patients were randomly assigned to a training set or a test set (3:1). An AutoML model was built from the training set, using 13 selected radiomic features from T2-weighted imaging, semantic features (ADC values), and clinical features (tumor length, tumor stage, lymph node metastasis status), and subsequent ten-fold cross-validation was performed. A test set was used to validate the proposed model. The AUC of the ROC curve was then calculated for the model. RESULTS: This AutoML model enabled robust differentiation of UC w/SD and pure UC in patients with MIBC in both training set (ten-fold cross-validation AUC = 0.955, 95% confidence interval [CI]: 0.944-0.965) and test set (AUC = 0.932, 95% CI: 0.812-1.000). CONCLUSION: The presented AutoML model, that incorporates the radiomic, semantic, and clinical features from baseline MRI, could be useful for preoperative differentiation of UC w/SD and pure UC. CLINICAL RELEVANCE STATEMENT: This MRI-based automated machine learning (AutoML) study provides a non-invasive and low-cost preoperative prediction tool to identify the muscle-invasive bladder cancer patients with variant histology, which may serve as a useful tool for clinical decision-making. KEY POINTS: • It is important to preoperatively diagnose variant histology from urothelial carcinoma in patients with muscle-invasive bladder carcinoma (MIBC), as their treatment strategy varies significantly. • An automated machine learning (AutoML) model based on baseline bladder MRI can identify the variant histology (squamous differentiation) from urothelial carcinoma preoperatively in patients with MIBC. • The developed AutoML model is a non-invasive and low-cost preoperative prediction tool, which may be useful for clinical decision-making.

Mol2Context-vec: learning molecular representation from context awareness for drug discovery.

With the rapid development of proteomics and the rapid increase of target molecules for drug action, computer-aided drug design (CADD) has become a basic task in drug discovery. One of the key challenges in CADD is molecular representation. High-quality molecular expression with chemical intuition helps to promote many boundary problems of drug discovery. At present, molecular representation still faces several urgent problems, such as the polysemy of substructures and unsmooth information flow between atomic groups. In this research, we propose a deep contextualized Bi-LSTM architecture, Mol2Context-vec, which can integrate different levels of internal states to bring dynamic representations of molecular substructures. And the obtained molecular context representation can capture the interactions between any atomic groups, especially a pair of atomic groups that are topologically distant. Experiments show that Mol2Context-vec achieves state-of-the-art performance on multiple benchmark datasets. In addition, the visual interpretation of Mol2Context-vec is very close to the structural properties of chemical molecules as understood by humans. These advantages indicate that Mol2Context-vec can be used as a reliable and effective tool for molecular expression. Availability: The source code is available for download in https://github.com/lol88/Mol2Context-vec.

The volume changes of unfolding of dsDNA.

High hydrostatic pressure can have profound effects on the stability of biomacromolecules. The magnitude and direction (stabilizing or destabilizing) of this effect is defined by the volume changes in the system, ΔV. Positive volume changes will stabilize the starting native state, whereas negative volume changes will lead to the stabilization of the final unfolded state. For the DNA double helix, experimental data suggested that when the thermostability of dsDNA is below 50°C, increase in hydrostatic pressure will lead to destabilization; i.e., helix-to-coil transition has negative ΔV. In contrast, the dsDNA sequences with the thermostability above 50°C showed positive ΔV values and were stabilized by hydrostatic pressure. In order to get insight into this switch in the response of dsDNA to hydrostatic pressure as a function of temperature, first we further validated this trend using experimental measurements of ΔV for 10 different dsDNA sequences using pressure perturbation calorimetry. We also developed a computational protocol to calculate the expected volume changes of dsDNA unfolding, which was benchmarked against the experimental set of 50 ΔV values that included, in addition to our data, the values from the literature. Computation predicts well the experimental values of ΔV. Such agreement between computation and experiment lends credibility to the computation protocol and provides molecular level rational for the observed temperature dependence of ΔV that can be traced to the hydration. Difference in the ΔV value for A/T versus G/C basepairs is also discussed.

Multitask deep learning with dynamic task balancing for quantum mechanical properties prediction.

Predicting quantum mechanical properties (QMPs) is very important for the innovation of material and chemistry science. Multitask deep learning models have been widely used in QMPs prediction. However, existing multitask learning models often train multiple QMPs prediction tasks simultaneously without considering the internal relationships and differences between tasks, which may cause the model to overfit easy tasks. In this study, we first proposed a multiscale dynamic attention graph neural network (MDGNN) for molecular representation learning. The MDGNN was designed in a multitask learning fashion that can solve multiple learning tasks at the same time. We then introduced a dynamic task balancing (DTB) strategy combining task differences and difficulties to reduce overfitting across multiple tasks. Finally, we adopted gradient-weighted class activation mapping (Grad-CAM) to analyze a deep learning model for frontier molecular orbital, highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy level predictions. We evaluated our approach using two large QMPs datasets and compared the proposed method to the state-of-the-art multitask learning models. The MDGNN outperforms other multitask learning approaches on two datasets. The DTB strategy can further improve the performance of MDGNN significantly. Moreover, we show that Grad-CAM creates explanations that are consistent with the molecular orbitals theory. These advantages demonstrate that the proposed method improves the generalization and interpretation capability of QMPs prediction modeling.

Predicting Drug-Target Interactions with Deep-Embedding Learning of Graphs and Sequences.

Computational approaches for predicting drug-target interactions (DTIs) play an important role in drug discovery since conventional screening experiments are time-consuming and expensive. In this study, we proposed end-to-end representation learning of a graph neural network with an attention mechanism and an attentive bidirectional long short-term memory (BiLSTM) to predict DTIs. For efficient training, we introduced a bidirectional encoder representations from transformers (BERT) pretrained method to extract substructure features from protein sequences and a local breadth-first search (BFS) to learn subgraph information from molecular graphs. Integrating both models, we developed a DTI prediction system. As a result, the proposed method achieved high performances with increases of 2.4% and 9.4% for AUC and recall, respectively, on unbalanced datasets compared with other methods. Extensive experiments showed that our model can relatively screen potential drugs for specific protein. Furthermore, visualizing the attention weights provides biological insight.

A novel artificial intelligence protocol to investigate potential leads for diabetes mellitus.

Dipeptidyl peptidase-4 (DPP4) is highly participated in regulating diabetes mellitus (DM), and inhibitors of DPP4 may act as potential DM drugs. Therefore, we performed a novel artificial intelligence (AI) protocol to screen and validate the potential inhibitors from Traditional Chinese Medicine Database. The potent top 10 compounds were selected as candidates by Dock Score. In order to further screen the candidates, we used numbers of machine learning regression models containing support vector machines, bagging, random forest and other regression algorithms, as well as deep neural network models to predict the activity of the candidates. In addition, as a traditional method, 2D QSAR (multiple linear regression) and 3D QSAR methods are also applied. The AI methods got a better performance than the traditional 2D QSAR method. Moreover, we also built a framework composed of deep neural networks and transformer to predict the binding affinity of candidates and DPP4. Artificial intelligence methods and QSAR models illustrated the compound, 2007_4105, was a potent inhibitor. The 2007_4105 compound was finally validated by molecular dynamics simulations. Combining all the models and algorithms constructed and the results, Hypecoum leptocarpum might be a potential and effective medicine herb for the treatment of DM.

Quantifying Athermality and Quantum Induced Deviations from Classical Fluctuation Relations.

In recent years, a quantum information theoretic framework has emerged for incorporating non-classical phenomena into fluctuation relations. Here, we elucidate this framework by exploring deviations from classical fluctuation relations resulting from the athermality of the initial thermal system and quantum coherence of the system's energy supply. In particular, we develop Crooks-like equalities for an oscillator system which is prepared either in photon added or photon subtracted thermal states and derive a Jarzynski-like equality for average work extraction. We use these equalities to discuss the extent to which adding or subtracting a photon increases the informational content of a state, thereby amplifying the suppression of free energy increasing process. We go on to derive a Crooks-like equality for an energy supply that is prepared in a pure binomial state, leading to a non-trivial contribution from energy and coherence on the resultant irreversibility. We show how the binomial state equality fits in relation to a previously derived coherent state equality and offers a richer feature-set.

Deep Learning and Random Forest Approach for Finding the Optimal Traditional Chinese Medicine Formula for Treatment of Alzheimer's Disease.

It has demonstrated that glycogen synthase kinase 3ß (GSK3ß) is related to Alzheimer's disease (AD). On the basis of the world largest traditional Chinese medicine (TCM) database, a network-pharmacology-based approach was utilized to investigate TCM candidates that can dock well with multiple targets. Support vector machine (SVM) and multiple linear regression (MLR) methods were utilized to obtain predicted models. In particular, the deep learning method and the random forest (RF) algorithm were adopted. We achieved R2 values of 0.927 on the training set and 0.862 on the test set with deep learning and 0.869 on the training set and 0.890 on the test set with RF. Besides, comparative molecular similarity indices analysis (CoMSIA) was performed to get a predicted model. All of the training models achieved good results on the test set. The stability of GSK3ß protein-ligand complexes was evaluated using 100 ns of MD simulation. Methyl 3- O-feruloylquinate and cynanogenin A induced both more compactness to the GSK3ß complex and stable conditions at all simulation times, and the GSK3ß complex also had no substantial fluctuations after a simulation time of 5 ns. For TCM molecules, we used the trained models to calculate predicted bioactivity values, and the optimum TCM candidates were obtained by ranking the predicted values. The results showed that methyl 3- O-feruloylquinate contained in Phellodendron amurense and cynanogenin A contained in Cynanchum atratum are capable of forming stable interactions with GSK3ß.

Phenylphthalazines as small-molecule inhibitors of urea transporter UT-B and their binding model.

AIM: Urea transporters (UT) are a family of transmembrane proteins that specifically transport urea. UT inhibitors exert diuretic activity without affecting electrolyte balance. The purpose of this study was to discover novel UT inhibitors and determine the inhibition mechanism. METHODS: The primary screening urea transporter B (UT-B) inhibitory activity was conducted in a collection of 10 000 diverse small molecules using mouse erythrocyte lysis assay. After discovering a hit with a core structure of 1-phenylamino-4-phenylphthalazin, the UT-B inhibitory activity of 160 analogs were examined with a stopped-flow light scattering assay and their structure-activity relationship (SAR) was analyzed. The inhibition mechanism was further investigated using in silico assays. RESULTS: A phenylphthalazine compound PU1424, chemically named 5-(4-((4-methoxyphenyl) amino) phthalazin-1-yl)-2-methylbenzene sulfonamide, showed potent UT-B inhibition activity, inhibited human and mouse UT-B-mediated urea transport with IC50 value of 0.02 and 0.69 µmol/L, respectively, and exerted 100% UT-B inhibition at higher concentrations. The compound PU1424 did not affect membrane urea transport in mouse erythrocytes lacking UT-B. Structure-activity analysis revealed that the analogs with methoxyl group at R4 and sulfonic amide at R2 position exhibited the highest potency inhibition activity on UT-B. Furthermore, in silico assays validated that the R4 and R2 positions of the analogs bound to the UT-B binding pocket and exerted inhibition activity on UT-B. CONCLUSION: The compound PU1424 is a novel inhibitor of both human and mouse UT-B with IC50 at submicromolar ranges. Its binding site is located at the So site of the UT-B structure.

ProteinVolume: calculating molecular van der Waals and void volumes in proteins.

BACKGROUND: Voids and cavities in the native protein structure determine the pressure unfolding of proteins. In addition, the volume changes due to the interaction of newly exposed atoms with solvent upon protein unfolding also contribute to the pressure unfolding of proteins. Quantitative understanding of these effects is important for predicting and designing proteins with predefined response to changes in hydrostatic pressure using computational approaches. The molecular surface volume is a useful metric that describes contribution of geometrical volume, which includes van der Waals volume and volume of the voids, to the total volume of a protein in solution, thus isolating the effects of hydration for separate calculations. RESULTS: We developed ProteinVolume, a highly robust and easy-to-use tool to compute geometric volumes of proteins. ProteinVolume generates the molecular surface of a protein and uses an innovative flood-fill algorithm to calculate the individual components of the molecular surface volume, van der Waals and intramolecular void volumes. ProteinVolume is user friendly and is available as a web-server or a platform-independent command-line version. CONCLUSIONS: ProteinVolume is a highly accurate and fast application to interrogate geometric volumes of proteins. ProteinVolume is a free web server available on http://gmlab.bio.rpi.edu . Free-standing platform-independent Java-based ProteinVolume executable is also freely available at this web site.

Methicillin resistance in Staphylococcus aureus requires glycosylated wall teichoic acids.

Staphylococcus aureus peptidoglycan (PG) is densely functionalized with anionic polymers called wall teichoic acids (WTAs). These polymers contain three tailoring modifications: d-alanylation, α-O-GlcNAcylation, and ß-O-GlcNAcylation. Here we describe the discovery and biochemical characterization of a unique glycosyltransferase, TarS, that attaches ß-O-GlcNAc (ß-O-N-acetyl-D-glucosamine) residues to S. aureus WTAs. We report that methicillin resistant S. aureus (MRSA) is sensitized to ß-lactams upon tarS deletion. Unlike strains completely lacking WTAs, which are also sensitive to ß-lactams, ΔtarS strains have no growth or cell division defects. Because neither α-O-GlcNAc nor ß-O-Glucose modifications can confer resistance, the resistance phenotype requires a highly specific chemical modification of the WTA backbone, ß-O-GlcNAc residues. These data suggest ß-O-GlcNAcylated WTAs scaffold factors required for MRSA resistance. The ß-O-GlcNAc transferase identified here, TarS, is a unique target for antimicrobials that sensitize MRSA to ß-lactams.

PSSP-MFFNet: A Multifeature Fusion Network for Protein Secondary Structure Prediction.

Protein secondary structure prediction (PSSP) is a fundamental task in modern bioinformatics research and is particularly important for uncovering the functional mechanisms of proteins. To improve the accuracy of PSSP, various general and essential features generated from amino acid sequences are often used for predicting the secondary structure. In this paper, we propose PSSP-MFFNet, a deep learning-based multi-feature fusion network for PSSP, which incorporates a multi-view deep learning architecture with the multiple sequence alignment (MSA) Transformer to efficiently capture global and local features of protein sequences. In practice, PSSP-MFFNet adopts a multi-feature fusion strategy, integrating different features generated from protein sequences, including MSA, sequence information, evolutionary information, and hidden state information. Moreover, we employ the MSA Transformer to interleave row and column attention across the input MSA. A hybrid network architecture of convolutional neural networks and long short-term memory networks is applied to extract high-level features after feature fusion. Furthermore, we introduce a transformer encoder to enhance the extracted high-level features. Comparative experimental results on independent tests demonstrate that PSSP-MFFNet has excellent generalization ability, outperforming other state-of-the-art PSSP models by an average of 1% on public benchmarks, including CASP12, CASP13, CASP14, TEST2018, and CB513. Our method can contribute to a better understanding of the biological functions of proteins, which has significant implications for drug discovery, disease diagnosis, and protein engineering.

MFTrans: A multi-feature transformer network for protein secondary structure prediction.

In the rapidly evolving field of computational biology, accurate prediction of protein secondary structures is crucial for understanding protein functions, facilitating drug discovery, and advancing disease diagnostics. In this paper, we propose MFTrans, a deep learning-based multi-feature fusion network aimed at enhancing the precision and efficiency of Protein Secondary Structure Prediction (PSSP). This model employs a Multiple Sequence Alignment (MSA) Transformer in combination with a multi-view deep learning architecture to effectively capture both global and local features of protein sequences. MFTrans integrates diverse features generated by protein sequences, including MSA, sequence information, evolutionary information, and hidden state information, using a multi-feature fusion strategy. The MSA Transformer is utilized to interleave row and column attention across the input MSA, while a Transformer encoder and decoder are introduced to enhance the extracted high-level features. A hybrid network architecture, combining a convolutional neural network with a bidirectional Gated Recurrent Unit (BiGRU) network, is used to further extract high-level features after feature fusion. In independent tests, our experimental results show that MFTrans has superior generalization ability, outperforming other state-of-the-art PSSP models by 3 % on average on public benchmarks including CASP12, CASP13, CASP14, TEST2016, TEST2018, and CB513. Case studies further highlight its advanced performance in predicting mutation sites. MFTrans contributes significantly to the protein science field, opening new avenues for drug discovery, disease diagnosis, and protein.

AI empowering traditional Chinese medicine?

For centuries, Traditional Chinese Medicine (TCM) has been a prominent treatment method in China, incorporating acupuncture, herbal remedies, massage, and dietary therapy to promote holistic health and healing. TCM has played a major role in drug discovery, with over 60% of small-molecule drugs approved by the FDA from 1981 to 2019 being derived from natural products. However, TCM modernization faces challenges such as data standardization and the complexity of TCM formulations. The establishment of comprehensive TCM databases has significantly improved the efficiency and accuracy of TCM research, enabling easier access to information on TCM ingredients and encouraging interdisciplinary collaborations. These databases have revolutionized TCM research, facilitating advancements in TCM modernization and patient care. In addition, advancements in AI algorithms and database data quality have accelerated progress in AI for TCM. The application of AI in TCM encompasses a wide range of areas, including herbal screening and new drug discovery, diagnostic and treatment principles, pharmacological mechanisms, network pharmacology, and the incorporation of innovative AI technologies. AI also has the potential to enable personalized medicine by identifying patterns and correlations in patient data, leading to more accurate diagnoses and tailored treatments. The potential benefits of AI for TCM are vast and diverse, promising continued progress and innovation in the field.

LDCNet: Lightweight dynamic convolution network for laparoscopic procedures image segmentation.

Medical image segmentation is fundamental for modern healthcare systems, especially for reducing the risk of surgery and treatment planning. Transanal total mesorectal excision (TaTME) has emerged as a recent focal point in laparoscopic research, representing a pivotal modality in the therapeutic arsenal for the treatment of colon & rectum cancers. Real-time instance segmentation of surgical imagery during TaTME procedures can serve as an invaluable tool in assisting surgeons, ultimately reducing surgical risks. The dynamic variations in size and shape of anatomical structures within intraoperative images pose a formidable challenge, rendering the precise instance segmentation of TaTME images a task of considerable complexity. Deep learning has exhibited its efficacy in Medical image segmentation. However, existing models have encountered challenges in concurrently achieving a satisfactory level of accuracy while maintaining manageable computational complexity in the context of TaTME data. To address this conundrum, we propose a lightweight dynamic convolution Network (LDCNet) that has the same superior segmentation performance as the state-of-the-art (SOTA) medical image segmentation network while running at the speed of the lightweight convolutional neural network. Experimental results demonstrate the promising performance of LDCNet, which consistently exceeds previous SOTA approaches. Codes are available at github.com/yinyiyang416/LDCNet.

MMFA-DTA: Multimodal Feature Attention Fusion Network for Drug-Target Affinity Prediction for Drug Repurposing Against SARS-CoV-2.

The continuous emergence of novel infectious diseases poses a significant threat to global public health security, necessitating the development of small-molecule inhibitors that directly target pathogens. The RNA-dependent RNA polymerase (RdRp) and main protease (Mpro) of SARS-CoV-2 have been validated as potential key antiviral drug targets for the treatment of COVID-19. However, the conventional new drug R&D cycle takes 10-15 years, failing to meet the urgent needs during epidemics. Here, we propose a general multimodal deep learning framework for drug repurposing, MMFA-DTA, to enable rapid virtual screening of known drugs and significantly improve discovery efficiency. By extracting graph topological and sequence features from both small molecules and proteins, we design attention mechanisms to achieve dynamic fusion across modalities. Results demonstrate the superior performance of MMFA-DTA in drug-target affinity prediction over several state-of-the-art baseline methods on Davis and KIBA data sets, validating the benefits of heterogeneous information integration for representation learning and interaction modeling. Further fine-tuning on COVID-19-relevant bioactivity data enhances model predictions for critical SARS-CoV-2 enzymes. Case studies screening the FDA-approved drug library successfully identify etacrynic acid as the potential lead compound against both RdRp and Mpro. Molecular dynamics simulations further confirm the stability and binding affinity of etacrynic acid to these targets. This study proves the great potential and advantages of deep learning and drug repurposing strategies in supporting antiviral drug discovery. The proposed general and rapid response computational framework holds significance for preparedness against future public health events.

Interaction-Based Inductive Bias in Graph Neural Networks: Enhancing Protein-Ligand Binding Affinity Predictions From 3D Structures.

Inductive bias in machine learning (ML) is the set of assumptions describing how a model makes predictions. Different ML-based methods for protein-ligand binding affinity (PLA) prediction have different inductive biases, leading to different levels of generalization capability and interpretability. Intuitively, the inductive bias of an ML-based model for PLA prediction should fit in with biological mechanisms relevant for binding to achieve good predictions with meaningful reasons. To this end, we propose an interaction-based inductive bias to restrict neural networks to functions relevant for binding with two assumptions: (1) A protein-ligand complex can be naturally expressed as a heterogeneous graph with covalent and non-covalent interactions; (2) The predicted PLA is the sum of pairwise atom-atom affinities determined by non-covalent interactions. The interaction-based inductive bias is embodied by an explainable heterogeneous interaction graph neural network (EHIGN) for explicitly modeling pairwise atom-atom interactions to predict PLA from 3D structures. Extensive experiments demonstrate that EHIGN achieves better generalization capability than other state-of-the-art ML-based baselines in PLA prediction and structure-based virtual screening. More importantly, comprehensive analyses of distance-affinity, pose-affinity, and substructure-affinity relations suggest that the interaction-based inductive bias can guide the model to learn atomic interactions that are consistent with physical reality. As a case study to demonstrate practical usefulness, our method is tested for predicting the efficacy of Nirmatrelvir against SARS-CoV-2 variants. EHIGN successfully recognizes the changes in the efficacy of Nirmatrelvir for different SARS-CoV-2 variants with meaningful reasons.