Protein kinase function and interactions with drugs are controlled in part by the movement of the DFG and ɑC-Helix motifs that are related to the catalytic activity of the kinase. Small molecule ligands elicit therapeutic effects with distinct selectivity profiles and residence times that often depend on the active or inactive kinase conformation(s) they bind. Modern AI-based structural modeling methods have the potential to expand upon the limited availability of experimentally determined kinase structures in inactive states. Here, we first explored the conformational space of kinases in the PDB and models generated by AlphaFold2 (AF2) and ESMFold, two prominent AI-based protein structure prediction methods. Our investigation of AF2's ability to explore the conformational diversity of the kinome at various multiple sequence alignment (MSA) depths showed a bias within the predicted structures of kinases in DFG-in conformations, particularly those controlled by the DFG motif, based on their overabundance in the PDB. We demonstrate that predicting kinase structures using AF2 at lower MSA depths explored these alternative conformations more extensively, including identifying previously unobserved conformations for 398 kinases. Ligand enrichment analyses for 23 kinases showed that, on average, docked models distinguished between active molecules and decoys better than random (average AUC (avgAUC) of 64.58), but select models perform well (e.g., avgAUCs for PTK2 and JAK2 were 79.28 and 80.16, respectively). Further analysis explained the ligand enrichment discrepancy between low- and high-performing kinase models as binding site occlusions that would preclude docking. The overall results of our analyses suggested that, although AF2 explored previously uncharted regions of the kinase conformational space and select models exhibited enrichment scores suitable for rational drug discovery, rigorous refinement of AF2 models is likely still necessary for drug discovery campaigns.

A continuing challenge in modern medicine is the identification of safer and more efficacious drugs. Precision therapeutics, which have one molecular target, have been long promised to be safer and more effective than traditional therapies. This approach has proven to be challenging for multiple reasons including lack of efficacy, rapidly acquired drug resistance, and narrow patient eligibility criteria. An alternative approach is the development of drugs that address the overall disease network by targeting multiple biological targets ('polypharmacology'). Rational development of these molecules will require improved methods for predicting single chemical structures that target multiple drug targets. To address this need, we developed the Multi-Targeting Drug DREAM Challenge, in which we challenged participants to predict single chemical entities that target pro-targets but avoid anti-targets for two unrelated diseases: RET-based tumors and a common form of inherited Tauopathy. Here, we report the results of this DREAM Challenge and the development of two neural network-based machine learning approaches that were applied to the challenge of rational polypharmacology. Together, these platforms provide a potentially useful first step towards developing lead therapeutic compounds that address disease complexity through rational polypharmacology.

Drosophila provides an inexpensive and quantitative platform for measuring whole animal drug response. A complementary approach is virtual screening, where chemical libraries can be efficiently screened against protein target(s). Here, we present a unique discovery platform integrating structure-based modeling with Drosophila biology and organic synthesis. We demonstrate this platform by developing chemicals targeting a Drosophila model of Medullary Thyroid Cancer (MTC) characterized by a transformation network activated by oncogenic dRetM955T. Structural models for kinases relevant to MTC were generated for virtual screening to identify unique preliminary hits that suppressed dRetM955T-induced transformation. We then combined features from our hits with those of known inhibitors to create a 'hybrid' molecule with improved suppression of dRetM955T transformation. Our platform provides a framework to efficiently explore novel kinase inhibitors outside of explored inhibitor chemical space that are effective in inhibiting cancer networks while minimizing whole body toxicity.

BACKGROUND: Genome-wide gene expression profiling of mammalian cells is becoming a staple of many published biomedical and biological research studies. Such data is deposited into data repositories such as the Gene Expression Omnibus (GEO) for potential reuse. However, these repositories currently do not provide simple interfaces to systematically analyze collections of related studies. RESULTS: Here we present GENE Expression and Enrichment Vector Analyzer (GEN3VA), a web-based system that enables the integrative analysis of aggregated collections of tagged gene expression signatures identified and extracted from GEO. Each tagged collection of signatures is presented in a report that consists of heatmaps of the differentially expressed genes; principal component analysis of all signatures; enrichment analysis with several gene set libraries across all signatures, which we term enrichment vector analysis; and global mapping of small molecules that are predicted to reverse or mimic each signature in the aggregate. We demonstrate how GEN3VA can be used to identify common molecular mechanisms of aging by analyzing tagged signatures from 244 studies that compared young vs. old tissues in mammalian systems. In a second case study, we collected 86 signatures from treatment of human cells with dexamethasone, a glucocorticoid receptor (GR) agonist. Our analysis confirms consensus GR target genes and predicts potential drug mimickers. CONCLUSIONS: GEN3VA can be used to identify, aggregate, and analyze themed collections of gene expression signatures from diverse but related studies. Such integrative analyses can be used to address concerns about data reproducibility, confirm results across labs, and discover new collective knowledge by data reuse. GEN3VA is an open-source web-based system that is freely available at: http://amp.pharm.mssm.edu/gen3va .

The Alanine-Serine-Cysteine transporter ASCT2 (SLC1A5) is a membrane protein that transports neutral amino acids into cells in exchange for outward movement of intracellular amino acids. ASCT2 is highly expressed in peripheral tissues such as the lung and intestines where it contributes to the homeostasis of intracellular concentrations of neutral amino acids. ASCT2 also plays an important role in the development of a variety of cancers such as melanoma by transporting amino acid nutrients such as glutamine into the proliferating tumors. Therefore, ASCT2 is a key drug target with potentially great pharmacological importance. Here, we identify seven ASCT2 ligands by computational modeling and experimental testing. In particular, we construct homology models based on crystallographic structures of the aspartate transporter GltPh in two different conformations. Optimization of the models' binding sites for protein-ligand complementarity reveals new putative pockets that can be targeted via structure-based drug design. Virtual screening of drugs, metabolites, fragments-like, and lead-like molecules from the ZINC database, followed by experimental testing of 14 top hits with functional measurements using electrophysiological methods reveals seven ligands, including five activators and two inhibitors. For example, aminooxetane-3-carboxylate is a more efficient activator than any other known ASCT2 natural or unnatural substrate. Furthermore, two of the hits inhibited ASCT2 mediated glutamine uptake and proliferation of a melanoma cancer cell line. Our results improve our understanding of how substrate specificity is determined in amino acid transporters, as well as provide novel scaffolds for developing chemical tools targeting ASCT2, an emerging therapeutic target for cancer and neurological disorders.

MOTIVATION: The mutation of amino acids often impacts protein function and structure. Mutations without negative effect sustain evolutionary pressure. We study a particular aspect of structural robustness with respect to mutations: regular protein secondary structure and natively unstructured (intrinsically disordered) regions. Is the formation of regular secondary structure an intrinsic feature of amino acid sequences, or is it a feature that is lost upon mutation and is maintained by evolution against the odds? Similarly, is disorder an intrinsic sequence feature or is it difficult to maintain? To tackle these questions, we in silico mutated native protein sequences into random sequence-like ensembles and monitored the change in predicted secondary structure and disorder. RESULTS: We established that by our coarse-grained measures for change, predictions and observations were similar, suggesting that our results were not biased by prediction mistakes. Changes in secondary structure and disorder predictions were linearly proportional to the change in sequence. Surprisingly, neither the content nor the length distribution for the predicted secondary structure changed substantially. Regions with long disorder behaved differently in that significantly fewer such regions were predicted after a few mutation steps. Our findings suggest that the formation of regular secondary structure is an intrinsic feature of random amino acid sequences, while the formation of long-disordered regions is not an intrinsic feature of proteins with disordered regions. Put differently, helices and strands appear to be maintained easily by evolution, whereas maintaining disordered regions appears difficult. Neutral mutations with respect to disorder are therefore very unlikely.

MOTIVATION: Natively unstructured (also dubbed intrinsically disordered) regions in proteins lack a defined 3D structure under physiological conditions and often adopt regular structures under particular conditions. Proteins with such regions are overly abundant in eukaryotes, they may increase functional complexity of organisms and they usually evade structure determination in the unbound form. Low propensity for the formation of internal residue contacts has been previously used to predict natively unstructured regions. RESULTS: We combined PROFcon predictions for protein-specific contacts with a generic pairwise potential to predict unstructured regions. This novel method, Ucon, outperformed the best available methods in predicting proteins with long unstructured regions. Furthermore, Ucon correctly identified cases missed by other methods. By computing the difference between predictions based on specific contacts (approach introduced here) and those based on generic potentials (realized in other methods), we might identify unstructured regions that are involved in protein-protein binding. We discussed one example to illustrate this ambitious aim. Overall, Ucon added quality and an orthogonal aspect that may help in the experimental study of unstructured regions in network hubs. AVAILABILITY: http://www.predictprotein.org/submit_ucon.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Natively unstructured or disordered protein regions may increase the functional complexity of an organism; they are particularly abundant in eukaryotes and often evade structure determination. Many computational methods predict unstructured regions by training on outliers in otherwise well-ordered structures. Here, we introduce an approach that uses a neural network in a very different and novel way. We hypothesize that very long contiguous segments with nonregular secondary structure (NORS regions) differ significantly from regular, well-structured loops, and that a method detecting such features could predict natively unstructured regions. Training our new method, NORSnet, on predicted information rather than on experimental data yielded three major advantages: it removed the overlap between testing and training, it systematically covered entire proteomes, and it explicitly focused on one particular aspect of unstructured regions with a simple structural interpretation, namely that they are loops. Our hypothesis was correct: well-structured and unstructured loops differ so substantially that NORSnet succeeded in their distinction. Benchmarks on previously used and new experimental data of unstructured regions revealed that NORSnet performed very well. Although it was not the best single prediction method, NORSnet was sufficiently accurate to flag unstructured regions in proteins that were previously not annotated. In one application, NORSnet revealed previously undetected unstructured regions in putative targets for structural genomics and may thereby contribute to increasing structural coverage of large eukaryotic families. NORSnet found unstructured regions more often in domain boundaries than expected at random. In another application, we estimated that 50%-70% of all worm proteins observed to have more than seven protein-protein interaction partners have unstructured regions. The comparative analysis between NORSnet and DISOPRED2 suggested that long unstructured loops are a major part of unstructured regions in molecular networks.

UNLABELLED: The mobility of a residue on the protein surface is closely linked to its function. The identification of extremely rigid or flexible surface residues can therefore contribute information crucial for solving the complex problem of identifying functionally important residues in proteins. Mobility is commonly measured by B-value data from high-resolution three-dimensional X-ray structures. Few methods predict B-values from sequence. Here, we present PROFbval, the first web server to predict normalized B-values from amino acid sequence. The server handles amino acid sequences (or alignments) as input and outputs normalized B-value and two-state (flexible/rigid) predictions. The server also assigns a reliability index for each prediction. For example, PROFbval correctly identifies residues in active sites on the surface of enzymes as particularly rigid. AVAILABILITY: http://www.rostlab.org/services/profbval CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.