HGTDR: Advancing drug repurposing with heterogeneous graph transformers.

MOTIVATION: Drug repurposing is a viable solution for reducing the time and cost associated with drug development. However, thus far, the proposed drug repurposing approaches still need to meet expectations. Therefore, it is crucial to offer a systematic approach for drug repurposing to achieve cost savings and enhance human lives. In recent years, using biological network-based methods for drug repurposing has generated promising results. Nevertheless, these methods have limitations. Primarily, the scope of these methods is generally limited concerning the size and variety of data they can effectively handle. Another issue arises from the treatment of heterogeneous data, which needs to be addressed or converted into homogeneous data, leading to a loss of information. A significant drawback is that most of these approaches lack end-to-end functionality, necessitating manual implementation and expert knowledge in certain stages. RESULTS: We propose a new solution, Heterogeneous Graph Transformer for Drug Repurposing (HGTDR), to address the challenges associated with drug repurposing. HGTDR is a three-step approach for knowledge graph-based drug repurposing: (1) constructing a heterogeneous knowledge graph, (2) utilizing a heterogeneous graph transformer network, and (3) computing relationship scores using a fully connected network. By leveraging HGTDR, users gain the ability to manipulate input graphs, extract information from diverse entities, and obtain their desired output. In the evaluation step, we demonstrate that HGTDR performs comparably to previous methods. Furthermore, we review medical studies to validate our method's top 10 drug repurposing suggestions, which have exhibited promising results. We also demonstrated HGTDR's capability to predict other types of relations through numerical and experimental validation, such as drug-protein and disease-protein inter-relations. AVAILABILITY AND IMPLEMENTATION: The source code and data are available at https://github.com/bcb-sut/HGTDR and http://git.dml.ir/BCB/HGTDR.

Integrin mechanosensing relies on a pivot-clip mechanism to reinforce cell adhesion.

Cells intricately sense mechanical forces from their surroundings, driving biophysical and biochemical activities. This mechanosensing phenomenon occurs at the cell-matrix interface, where mechanical forces resulting from cellular motion, such as migration or matrix stretching, are exchanged through surface receptors, primarily integrins, and their corresponding matrix ligands. A pivotal player in this interaction is the α5ß1 integrin and fibronectin (FN) bond, known for its role in establishing cell adhesion sites for migration. However, upregulation of the α5ß1-FN bond is associated with uncontrolled cell metastasis. This bond operates through catch bond dynamics, wherein the bond lifetime paradoxically increases with greater force. The mechanism sustaining the characteristic catch bond dynamics of α5ß1-FN remains unclear. Leveraging molecular dynamics simulations, our approach unveils a pivot-clip mechanism. Two key binding sites on FN, namely the synergy site and the RGD (Arg-Gly-Asp) motif, act as active points for structural changes in α5ß1 integrin. Conformational adaptations at these sites are induced by a series of hydrogen bond formations and breaks at the synergy site. We disrupt these adaptations through a double mutation on FN, known to reduce cell adhesion. A whole-cell finite-element model is employed to elucidate how the synergy site may promote dynamic α5ß1-FN binding, resisting cell contraction. In summary, our study integrates molecular- and cellular-level modeling to propose that FN's synergy site reinforces cell adhesion through enhanced binding dynamics and a mechanosensitive pivot-clip mechanism. This work sheds light on the interplay between mechanical forces and cell-matrix interactions, contributing to our understanding of cellular behaviors in physiological and pathological contexts.

GO Bench: shared hub for universal benchmarking of machine learning-based protein functional annotations.

MOTIVATION: Gene annotation is the problem of mapping proteins to their functions represented as Gene Ontology (GO) terms, typically inferred based on the primary sequences. Gene annotation is a multi-label multi-class classification problem, which has generated growing interest for its uses in the characterization of millions of proteins with unknown functions. However, there is no standard GO dataset used for benchmarking the newly developed new machine learning models within the bioinformatics community. Thus, the significance of improvements for these models remains unclear. RESULTS: The Gene Benchmarking database is the first effort to provide an easy-to-use and configurable hub for the learning and evaluation of gene annotation models. It provides easy access to pre-specified datasets and takes the non-trivial steps of preprocessing and filtering all data according to custom presets using a web interface. The GO bench web application can also be used to evaluate and display any trained model on leaderboards for annotation tasks. AVAILABILITY AND IMPLEMENTATION: The GO Benchmarking dataset is freely available at www.gobench.org. Code is hosted at github.com/mofradlab, with repositories for website code, core utilities and examples of usage (Supplementary Section S.7). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Exploring the performance and explainability of fine-tuned BERT models for neuroradiology protocol assignment.

BACKGROUND: Deep learning has demonstrated significant advancements across various domains. However, its implementation in specialized areas, such as medical settings, remains approached with caution. In these high-stake environments, understanding the model's decision-making process is critical. This study assesses the performance of different pretrained Bidirectional Encoder Representations from Transformers (BERT) models and delves into understanding its decision-making within the context of medical image protocol assignment. METHODS: Four different pre-trained BERT models (BERT, BioBERT, ClinicalBERT, RoBERTa) were fine-tuned for the medical image protocol classification task. Word importance was measured by attributing the classification output to every word using a gradient-based method. Subsequently, a trained radiologist reviewed the resulting word importance scores to assess the model's decision-making process relative to human reasoning. RESULTS: The BERT model came close to human performance on our test set. The BERT model successfully identified relevant words indicative of the target protocol. Analysis of important words in misclassifications revealed potential systematic errors in the model. CONCLUSIONS: The BERT model shows promise in medical image protocol assignment by reaching near human level performance and identifying key words effectively. The detection of systematic errors paves the way for further refinements to enhance its safety and utility in clinical settings.

Force transmission and SUN-KASH higher-order assembly in the LINC complex models.

The linkers of the nucleoskeleton and cytoskeleton (LINC) complex comprises Sad-1 and UNC-84 (SUN) and Klarsicht, ANC-1, SYNE homology (KASH) domain proteins, whose conserved interactions provide a physical coupling between the cytoskeleton and the nucleoskeleton, thereby mediating the transfer of physical forces across the nuclear envelope. The LINC complex can perform distinct cellular functions by pairing various KASH domain proteins with the same SUN domain protein. Recent studies have suggested a higher-order assembly of SUN and KASH instead of a more widely accepted linear trimer model for the LINC complex. In the present study, we use molecular dynamics simulations to investigate the mechanism of force transfer across the two proposed models of LINC complex assembly, namely the 3:3 linear trimer model and the 6:6 higher-order model. Employing steered molecular dynamics simulations with various structures using forces at different rates and directions, we examine the structural stability of the two models under various biologically relevant conditions. Our results suggest that both models can withstand and transfer significant levels of force while retaining their structural integrity. However, the force response of various SUN/KASH assemblies depend on the force direction and pulling rates. Slower pulling rates result in higher mean square fluctuations of the 3:3 assembly compared to the fast pulling. Interestingly, the 6:6 assembly tends to provide an additional range of motion flexibility and might be more advantageous to the structural rigidity and pliability of the nuclear envelope. These findings offer insights into how the SUN and KASH proteins maintain the structural integrity of the nuclear membrane.

Molecular models of LINC complex assembly at the nuclear envelope.

Large protein complexes assemble at the nuclear envelope to transmit mechanical signals between the cytoskeleton and nucleoskeleton. These protein complexes are known as the linkers of the nucleoskeleton and cytoskeleton complexes (LINC complexes) and are formed by the interaction of SUN and KASH domain proteins in the nuclear envelope. Ample evidence suggests that SUN-KASH complexes form higher-order assemblies to withstand and transfer forces across the nuclear envelope. Herein, we present a review of recent studies over the past few years that have shed light on the mechanisms of SUN-KASH interactions, their higher order assembly, and the molecular mechanisms of force transfer across these complexes.

On the Sensitivity and Affinity of Gold, Silver, and Platinum Surfaces against the SARS-CoV-2 Virus: A Comparative Computational Study.

The novel coronavirus disease and its complications have motivated the design of new sensors with the highest sensitivity, and affinity for the detection of the SARS-CoV-2 virus is considered in many research studies. In this research article, we employ full atomistic molecular dynamics (MD) models to study the interactions between the receptor binding domain (RBD) and spike protein of the coronavirus and different metals such as gold (Au), platinum (Pt), and silver (Ag) to analyze their sensitivity against this virus. The comparison between the RBD interactions with ACE2 (angiotensin-converting enzyme 2) and different metals indicates that metals have remarkable effects on the structural features and dynamical properties of the RBD. The binding site of the RBD has more affinity to the surfaces of gold, platinum, and silver than to the other parts of the protein. Moreover, the initial configuration of the RBD relative to the metal surface plays an important role in the stability of metal complexes with the RBD. The binding face of the protein to the metal surface has been changed in the presence of different metals. In other words, the residues of the RBD that participate in RBD interactions with the metals are different irrespective of the initial configurations in which the [Asn, Thr, Tyr], [Ser, Thr, Tyr], and [Asn, Asp, Tyr] residues of the protein have a greater affinity to Ag, Au, and Pt, respectively. The corresponding metals have a considerable affinity to the RBD, which due to strong interactions with the protein can change the secondary structure and structural features. Based on the obtained results during the complexation process between the protein and metals, the helical structure of the protein changes to the bend and antiparallel ß-sheets. The calculated binding energies for the RBD complexes with silver, gold, and platinum are -95.03, -138.03, and -133.96 kcal·mol-1, respectively. The adsorption process of the spike protein on the surfaces of different metals represents similar results and indicates that the entire spike protein of the coronavirus forms a more stable complex with the gold surface compared with other metals. Moreover, the RBD of the spike protein has more interactions with the surfaces than with the other parts of the protein. Therefore, it is possible to predict the properties of the coronavirus on the metal surface based on the dynamical behavior of the RBD. Overall, our computational results confirm that the gold surface can be considered as an outstanding substrate for developing new sensors with the highest sensitivity against SARS-CoV-2.

Methylation at a conserved lysine residue modulates tau assembly and cellular functions.

The neuronal microtubule-associated protein tau undergoes multiple post-translational modifications, which dynamically modulate its molecular functions and biochemical features in space and time. Among them, we have recently reported that a conserved lysine residue mapping to the microtubule-binding domain of the protein (K306 in mouse and K317 in human) is differentially methylated in a model of chronic autoimmune demyelination. In contrast with other well-studied tau post-translational modifications such as phosphorylation, lysine methylation is far less investigated and its specific impact on tau biology is not fully understood. Here we performed a comprehensive analysis of the effects of K317 methylation on key tau features. By combining in silico simulations with in vitro biochemical assays and live-cell imaging, we show that methylated tau is more prone to self-assembly into insoluble structures. Moreover, we demonstrate that K317 methylation affects the stabilization activity of tau on microtubule dynamics. Lastly, we highlight a role for K317 methylation in regulating both neuronal differentiation and cell proliferation. Altogether, these findings shed light on the biology of an overlooked tau post-translational modification as well as on the fine tuning of tau functionality in health and disease.

EpitopeVec: linear epitope prediction using deep protein sequence embeddings.

MOTIVATION: B-cell epitopes (BCEs) play a pivotal role in the development of peptide vaccines, immuno-diagnostic reagents and antibody production, and thus in infectious disease prevention and diagnostics in general. Experimental methods used to determine BCEs are costly and time-consuming. Therefore, it is essential to develop computational methods for the rapid identification of BCEs. Although several computational methods have been developed for this task, generalizability is still a major concern, where cross-testing of the classifiers trained and tested on different datasets has revealed accuracies of 51-53%. RESULTS: We describe a new method called EpitopeVec, which uses a combination of residue properties, modified antigenicity scales, and protein language model-based representations (protein vectors) as features of peptides for linear BCE predictions. Extensive benchmarking of EpitopeVec and other state-of-the-art methods for linear BCE prediction on several large and small datasets, as well as cross-testing, demonstrated an improvement in the performance of EpitopeVec over other methods in terms of accuracy and area under the curve. As the predictive performance depended on the species origin of the respective antigens (viral, bacterial and eukaryotic), we also trained our method on a large viral dataset to create a dedicated linear viral BCE predictor with improved cross-testing performance. AVAILABILITY AND IMPLEMENTATION: The software is available at https://github.com/hzi-bifo/epitope-prediction. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

A short HLA-DRA isoform binds the HLA-DR2 heterodimer on the outer domain of the peptide-binding site.

The human leukocyte antigen (HLA) locus encodes a large group of proteins governing adaptive and innate immune responses. Among them, HLA class II proteins form α/ß heterodimers on the membrane of professional antigen-presenting cells (APCs), where they display both, self and pathogen-derived exogenous antigens to CD4+ T lymphocytes. We have previously shown that a shorter HLA-DRA isoform (sHLA-DRA) lacking 25 amino acids can be presented onto the cell membrane via binding to canonical HLA-DR2 heterodimers. Here, we employed atomistic molecular dynamics simulations to decipher the binding position of sHLA-DRA and its structural impact on functional regions of the HLA-DR2 molecule. We show that a loop region exposed only in the short isoform (residues R69 to G83) is responsible for binding to the outer domain of the HLA-DR2 peptide-binding site, and experimentally validated the critical role of F76 in mediating such interaction. Additionally, sHLA-DRA allosterically modifies the peptide-binding pocket conformation. In summary, this study unravels key molecular mechanisms underlying sHLA-DRA function, providing important insights into the role of full-length proteins in structural modulation of HLA class II receptors.

Acid-Sensitive Surfactants Enhance the Delivery of Nucleic Acids.

The development of endosomal disruptive agents is a major challenge in the field of drug delivery and pharmaceutical chemistry. Current endosomal disruptive agents are composed of polymers, peptides, and nanoparticles and have had limited clinical impact. Alternatives to traditional endosomal disruptive agents are therefore greatly needed. In this report, we introduce a new class of low molecular weight endosomal disruptive agents, termed caged surfactants, that selectively disrupt endosomes via reversible PEGylation under acidic endosomal conditions. The caged surfactants have the potential to address several of the limitations hindering the development of current endosomal disruptive agents, such as high toxicity and low excretion, and are amenable to traditional medicinal chemistry approaches for optimization. In this report, we synthesized three generations of caged surfactants and demonstrated that they can enhance the ability of cationic lipids to deliver mRNA into primary cells. We also show that caged surfactants can deliver siRNA into cells when modified with the RNA-binding dye thiazole orange. We anticipate that the caged surfactants will have numerous applications in pharmaceutical chemistry and drug delivery given their versatility.

Killer Cell Immunoglobulin-like Receptor Variants Are Associated with Protection from Symptoms Associated with More Severe Course in Parkinson Disease.

Immune dysfunction plays a role in the development of Parkinson disease (PD). NK cells regulate immune functions and are modulated by killer cell immunoglobulin-like receptors (KIR). KIR are expressed on the surface of NK cells and interact with HLA class I ligands on the surface of all nucleated cells. We investigated KIR-allelic polymorphism to interrogate the role of NK cells in PD. We sequenced KIR genes from 1314 PD patients and 1978 controls using next-generation methods and identified KIR genotypes using custom bioinformatics. We examined associations of KIR with PD susceptibility and disease features, including age at disease onset and clinical symptoms. We identified two KIR3DL1 alleles encoding highly expressed inhibitory receptors associated with protection from PD clinical features in the presence of their cognate ligand: KIR3DL1*015/HLA-Bw4 from rigidity (p c = 0.02, odds ratio [OR] = 0.39, 95% confidence interval [CI] 0.23-0.69) and KIR3DL1*002/HLA-Bw4i from gait difficulties (p c = 0.05, OR = 0.62, 95% CI 0.44-0.88), as well as composite symptoms associated with more severe disease. We also developed a KIR3DL1/HLA interaction strength metric and found that weak KIR3DL1/HLA interactions were associated with rigidity (pc = 0.05, OR = 9.73, 95% CI 2.13-172.5). Highly expressed KIR3DL1 variants protect against more debilitating symptoms of PD, strongly implying a role of NK cells in PD progression and manifestation.

Decoding the Effect of Hydrostatic Pressure on TRPV1 Lower-Gate Conformation by Molecular-Dynamics Simulation.

In response to hydrostatic pressure, the cation channel transient receptor potential vanilloid 1 (TRPV1) is essential in signaling pathways linked to glaucoma. When activated, TRPV1 undergoes a gating transition from a closed to an open state that allows the influx of Ca2+ ions. However, the gating mechanism of TRPV1 in response to hydrostatic pressure at the molecular level is still lacking. To understand the effect of hydrostatic pressure on the activation of TRPV1, we conducted molecular-dynamics (MD) simulations on TRPV1 under different hydrostatic pressure configurations, with and without a cell membrane. The TRPV1 membrane-embedded model is more stable than the TPRV1-only model, indicating the importance of including the cell membrane in MD simulation. Under elevated pressure at 27.6 mmHg, we observed a more dynamic and outward motion of the TRPV1 domains in the lower-gate area than in the simulation under normal pressure at 12.6 mmHg. While a complete closed-to-open-gate transition was not evident in the limited course of our MD simulations, an increase in the channel radius at the lower gate was observed at 27.6 mmHg versus that at 12.6 mmHg. These findings provide novel information regarding the effect of hydrostatic pressure on TRPV1 channels.

Free energy calculations shed light on the nuclear pore complex's selective barrier nature.

The nuclear pore complex (NPC) is the exclusive gateway for traffic control across the nuclear envelope. Although smaller cargoes (less than 5-9 nm in size) can freely diffuse through the NPC, the passage of larger cargoes is restricted to those accompanied by nuclear transport receptors (NTRs). This selective barrier nature of the NPC is putatively associated with the intrinsically disordered, phenylalanine-glycine repeat-domains containing nucleoporins, termed FG-Nups. The precise mechanism underlying how FG-Nups carry out such an exquisite task at high throughputs has, however, remained elusive and the subject of various hypotheses. From the thermodynamics perspective, free energy analysis can be a way to determine cargo's transportability because the traffic through the NPC must be in the direction of reducing the free energy. In this study, we developed a computational model to evaluate the free energy composed of the conformational entropy of FG-Nups and the energetic gain associated with binding interactions between FG-Nups and NTRs and investigated whether these physical features can be the basis of NPC's selectivity. Our results showed that the reduction in conformational entropy by inserting a cargo into the NPC increased the free energy by an amount substantially greater than the thermal energy (≫kBT), whereas the free energy change was negligible (

FG nucleoporins feature unique patterns that distinguish them from other IDPs.

FG nucleoporins (FG Nups) are intrinsically disordered proteins and are the putative regulators of nucleocytoplasmic transport. They allow fast, yet selective, transport of molecules through the nuclear pore complex, but the underlying mechanism of nucleocytoplasmic transport is not yet fully discovered. As a result, FG Nups have been the subject of extensive research in the past two decades. Although most studies have been focused on analyzing the conformation and function of FG Nups from a biophysical standpoint, some recent studies have investigated the sequence-function relationship of FG Nups, with a few investigating amino acid sequences of a large number of FG Nups to understand common characteristics that might enable their function. Previously, we identified an evolutionarily conserved feature in FG Nup sequences, which are extended subsequences with low charge density, containing only positive charges, and located toward the N-terminus of FG Nups. We named these patterns longest positive like charge regions (lpLCRs). These patterns are specific to positively charged residues, and negatively charged residues do not demonstrate such a pattern. In this study, we compare FG Nups with other disordered proteins obtained from the DisProt and UniProt database in terms of presence of lpLCRs. Our results show that the lpLCRs are virtually exclusive to FG Nups and are not observed in other disordered proteins. Also, lpLCRs are what differentiate FG Nups from DisProt proteins in terms of charge distribution, meaning that excluding lpLCRs from the sequences of FG Nups make them similar to DisProt proteins in terms of charge distribution. We also previously showed the biophysical effect of lpLCRs in conformation of FG Nups. The results of this study are in line with our previous findings and imply that lpLCRs are virtually exclusive and functionally significant characteristics of FG Nups and nucleocytoplasmic transport.

A splice acceptor variant in HLA-DRA affects the conformation and cellular localization of the class II DR alpha-chain.

Class II human leucocyte antigen (HLA) proteins are involved in the immune response by presenting pathogen-derived peptides to CD4+ T lymphocytes. At the molecular level, they are constituted by α/ß-heterodimers on the surface of professional antigen-presenting cells. Here, we report that the acceptor variant (rs8084) in the HLA-DRA gene mediates the transcription of an alternative version of the α-chain lacking 25 amino acids in its extracellular domain. Molecular dynamics simulations suggest this isoform undergoes structural refolding which in turn affects its stability and cellular trafficking. The short HLA-DRA isoform cannot reach the cell surface, although it is still able to bind the corresponding ß-chain. Conversely, it remains entrapped within the endoplasmic reticulum where it is targeted for degradation. Furthermore, we demonstrate that the short isoform can be transported to the cell membrane via interactions with the peptide-binding site of canonical HLA heterodimers. Altogether, our findings indicate that short HLA-DRA functions as a novel intact antigen for class II HLA molecules.

Characterizing Binding Interactions That Are Essential for Selective Transport through the Nuclear Pore Complex.

Specific macromolecules are rapidly transported across the nuclear envelope via the nuclear pore complex (NPC). The selective transport process is facilitated when nuclear transport receptors (NTRs) weakly and transiently bind to intrinsically disordered constituents of the NPC, FG Nups. These two types of proteins help maintain the selective NPC barrier. To interrogate their binding interactions in vitro, we deployed an NPC barrier mimic. We created the stationary phase by covalently attaching fragments of a yeast FG Nup called Nsp1 to glass coverslips. We used a tunable mobile phase containing NTR, nuclear transport factor 2 (NTF2). In the stationary phase, three main factors affected binding: the number of FG repeats, the charge of fragments, and the fragment density. We also identified three main factors affecting binding in the mobile phase: the avidity of the NTF2 variant for Nsp1, the presence of nonspecific proteins, and the presence of additional NTRs. We used both experimentally determined binding parameters and molecular dynamics simulations of Nsp1FG fragments to create an agent-based model. The results suggest that NTF2 binding is negatively cooperative and dependent on the density of Nsp1FG molecules. Our results demonstrate the strengths of combining experimental and physical modeling approaches to study NPC-mediated transport.

Kindlin Assists Talin to Promote Integrin Activation.

Integrin αIIbß3 is a predominant type of integrin abundantly expressed on the surface of platelets and its activation regulates the process of thrombosis. Talin and kindlin are cytoplasmic proteins that bind to integrin and modulate its affinity for extracellular ligands. Although the molecular details of talin-mediated integrin activation are known, the mechanism of kindlin involvement in this process remains elusive. Here, we demonstrate that the interplay between talin and kindlin promotes integrin activation. Our all-atomic molecular dynamics simulations on complete transmembrane and cytoplasmic domains of integrin αIIbß3, talin1 F2/F3 subdomains, and the kindlin2 FERM domain in an explicit lipid-water environment over a microsecond timescale unraveled the role of kindlin as an enhancer of the talin interaction with the membrane proximal region of ß-integrin. The cooperation of kindlin with talin results in a complete disruption of salt bridges between R995 on αIIb and D723/E726 on ß3. Furthermore, kindlin modifies the molecular mechanisms of inside-out activation by decreasing the crossing angle between transmembrane helices of integrin αIIbß3, which eventually results in parallelization of integrin dimer. In addition, our control simulation featuring integrin in complex with kindlin reveals that kindlin binding is not sufficient for unclasping the inner-membrane and outer-membrane interactions of integrin dimer, thus ruling out the possibility of solitary action of kindlin in integrin activation.

DiTaxa: nucleotide-pair encoding of 16S rRNA for host phenotype and biomarker detection.

SUMMARY: Identifying distinctive taxa for micro-biome-related diseases is considered key to the establishment of diagnosis and therapy options in precision medicine and imposes high demands on the accuracy of micro-biome analysis techniques. We propose an alignment- and reference- free subsequence based 16S rRNA data analysis, as a new paradigm for micro-biome phenotype and biomarker detection. Our method, called DiTaxa, substitutes standard operational taxonomic unit (OTU)-clustering by segmenting 16S rRNA reads into the most frequent variable-length subsequences. We compared the performance of DiTaxa to the state-of-the-art methods in phenotype and biomarker detection, using human-associated 16S rRNA samples for periodontal disease, rheumatoid arthritis and inflammatory bowel diseases, as well as a synthetic benchmark dataset. DiTaxa performed competitively to the k-mer based state-of-the-art approach in phenotype prediction while outperforming the OTU-based state-of-the-art approach in finding biomarkers in both resolution and coverage evaluated over known links from literature and synthetic benchmark datasets. AVAILABILITY AND IMPLEMENTATION: DiTaxa is available under the Apache 2 license at http://llp.berkeley.edu/ditaxa. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Agent-Based Modeling in Molecular Systems Biology.

Molecular systems orchestrating the biology of the cell typically involve a complex web of interactions among various components and span a vast range of spatial and temporal scales. Computational methods have advanced our understanding of the behavior of molecular systems by enabling us to test assumptions and hypotheses, explore the effect of different parameters on the outcome, and eventually guide experiments. While several different mathematical and computational methods are developed to study molecular systems at different spatiotemporal scales, there is still a need for methods that bridge the gap between spatially-detailed and computationally-efficient approaches. In this review, we summarize the capabilities of agent-based modeling (ABM) as an emerging molecular systems biology technique that provides researchers with a new tool in exploring the dynamics of molecular systems/pathways in health and disease.