ModFlex: Towards Function Focused Protein Modeling.
Ontology highlight
ABSTRACT:
SUBMITTER: Sedova M
PROVIDER: S-EPMC8428579 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Modeling the dynamic nature of protein-ligand binding with atomistic simulations is one of the main challenges in computational biophysics, with important implications in the drug design process. Although in the past few years hardware and software advances have significantly revamped the use of molecular simulations, we still lack a fast and accurate ab initio description of the binding mechanism in complex systems, available only for up-to-date techniques and requiring several hours or days of heavy computation. Such delay is one of the main limiting factors for a larger penetration of protein dynamics modeling in the pharmaceutical industry. Here we present a game-changing technology, opening up the way for fast reliable simulations of protein dynamics by combining an adaptive reinforcement learning procedure with Monte Carlo sampling in the frame of modern multi-core computational resources. We show remarkable performance in mapping the protein-ligand energy landscape, being able to reproduce the full binding mechanism in less than half an hour, or the active site induced fit in less than 5?minutes. We exemplify our method by studying diverse complex targets, including nuclear hormone receptors and GPCRs, demonstrating the potential of using the new adaptive technique in screening and lead optimization studies.
| S-EPMC5559483 | biostudies-literature
Project description:Gene Ontology (GO) is the largest resource for cataloging gene products. This resource grows steadily and, naturally, this growth raises issues regarding the structure of the ontology. Moreover, modeling and refactoring large ontologies such as GO is generally far from being simple, as a whole as well as when focusing on certain aspects or fragments. It seems that human-friendly graphical modeling languages such as the Unified Modeling Language (UML) could be helpful in connection with these tasks.We investigate the use of UML for making the structural organization of the Molecular Function Ontology (MFO), a sub-ontology of GO, more explicit. More precisely, we present a UML dialect, called the Function Modeling Language (FueL), which is suited for capturing functions in an ontologically founded way. FueL is equipped, among other features, with language elements that arise from studying patterns of subsumption between functions. We show how to use this UML dialect for capturing the structure of molecular functions. Furthermore, we propose and discuss some refactoring options concerning fragments of MFO.FueL enables the systematic, graphical representation of functions and their interrelations, including making information explicit that is currently either implicit in MFO or is mainly captured in textual descriptions. Moreover, the considered subsumption patterns lend themselves to the methodical analysis of refactoring options with respect to MFO. On this basis we argue that the approach can increase the comprehensibility of the structure of MFO for humans and can support communication, for example, during revision and further development.
| S-EPMC5628467 | biostudies-literature
Project description:Synthetic antibody libraries have proven immensely useful for the de novo isolation of antibodies without the need for animal immunization. Recently, focused libraries designed to recognize particular classes of ligands, such as haptens or proteins, have been employed to facilitate the selection of high-affinity antibodies. Focused libraries are built using V regions encoding combinations of canonical structures that resemble the structural features of antibodies that bind the desired class of ligands and sequence diversity is introduced at residues typically involved in recognition. Here we describe the generation and experimental validation of two different single-chain antibody variable fragment libraries that efficiently generate binders to peptides, a class of molecules that has proven to be a difficult target for antibody generation. First, a human anti-peptide library was constructed by diversifying a scaffold: the human variable heavy chain (V(H)) germ line gene 3-23, which was fused to a variant of the human variable light chain (V(L)) germ line gene A27, in which L1 was modified to encode the canonical structure found in anti-peptide antibodies. The sequence diversity was introduced into 3-23 (V(H)) only, targeting for diversification residues commonly found in contact with protein and peptide antigens. Second, a murine library was generated using the antibody 26-10, which was initially isolated based on its affinity to the hapten digoxin, but also binds peptides and exhibits a canonical structure pattern typical of anti-peptide antibodies. Diversity was introduced in the V(H) only using the profile of amino acids found at positions that frequently contact peptide antigens. Both libraries yielded binders to two model peptides, angiotensin and neuropeptide Y, following screening by solution phage panning. The mouse library yielded antibodies with affinities below 20 nM to both targets, although only the V(H) had been subjected to diversification.
| S-EPMC2494707 | biostudies-literature
Project description:Protein rigidity analysis is an efficient computational method for extracting flexibility information from static, X-ray crystallography protein data. Atoms and bonds are modeled as a mechanical structure and analyzed with a fast graph-based algorithm, producing a decomposition of the flexible molecule into interconnected rigid clusters. The result depends critically on noncovalent atomic interactions, primarily on how hydrogen bonds and hydrophobic interactions are computed and modeled. Ongoing research points to the stringent need for benchmarking rigidity analysis software systems, towards the goal of increasing their accuracy and validating their results, either against each other and against biologically relevant (functional) parameters. We propose two new methods for modeling hydrogen bonds and hydrophobic interactions that more accurately reflect a mechanical model, without being computationally more intensive. We evaluate them using a novel scoring method, based on the B-cubed score from the information retrieval literature, which measures how well two cluster decompositions match.To evaluate the modeling accuracy of KINARI, our pebble-game rigidity analysis system, we use a benchmark data set of 20 proteins, each with multiple distinct conformations deposited in the Protein Data Bank. Cluster decompositions for them were previously determined with the RigidFinder method from Gerstein's lab and validated against experimental data. When KINARI's default tuning parameters are used, an improvement of the B-cubed score over a crude baseline is observed in 30% of this data. With our new modeling options, improvements were observed in over 70% of the proteins in this data set. We investigate the sensitivity of the cluster decomposition score with case studies on pyruvate phosphate dikinase and calmodulin.To substantially improve the accuracy of protein rigidity analysis systems, thorough benchmarking must be performed on all current systems and future extensions. We have measured the gain in performance by comparing different modeling methods for noncovalent interactions. We showed that new criteria for modeling hydrogen bonds and hydrophobic interactions can significantly improve the results. The two new methods proposed here have been implemented and made publicly available in the current version of KINARI (v1.3), together with the benchmarking tools, which can be downloaded from our software's website, http://kinari.cs.umass.edu.
| S-EPMC3817810 | biostudies-literature
Project description:Deep machine learning is expanding the conceptual framework and capacity of computational compound design, enabling new applications through generative modeling. We have explored the systematic design of covalent protein kinase inhibitors by learning from kinome-relevant chemical space, followed by focusing on an exemplary kinase of interest. Covalent inhibitors experience a renaissance in drug discovery, especially for targeting protein kinases. However, computational design of this class of inhibitors has thus far only been little investigated. To this end, we have devised a computational approach combining fragment-based design and deep generative modeling augmented by three-dimensional pharmacophore screening. This approach is thought to be particularly relevant for medicinal chemistry applications because it combines knowledge-based elements with deep learning and is chemically intuitive. As an exemplary application, we report for Bruton's tyrosine kinase (BTK), a major drug target for the treatment of inflammatory diseases and leukemia, the generation of novel candidate inhibitors with a specific chemically reactive group for covalent modification, requiring only little target-specific compound information to guide the design efforts. Newly generated compounds include known inhibitors and characteristic substructures and many novel candidates, thus lending credence to the computational approach, which is readily applicable to other targets.
| S-EPMC8778003 | biostudies-literature
Project description:In the current study, we set out to examine the viability of a novel approach to modeling human personality. Research in psychology suggests that people's personalities can be effectively described using five broad dimensions (the Five-Factor Model; FFM); however, the FFM potentially leaves room for improved predictive accuracy. We propose a novel approach to modeling human personality that is based on the maximization of the model's predictive accuracy. Unlike the FFM, which performs unsupervised dimensionality reduction, we utilized a supervised machine learning technique for dimensionality reduction of questionnaire data, using numerous psychologically meaningful outcomes as data labels (e.g., intelligence, well-being, sociability). The results showed that our five-dimensional personality summary, which we term the "Predictive Five" (PF), provides predictive performance that is better than the FFM on two independent validation datasets, and on a new set of outcome variables selected by an independent group of psychologists. The approach described herein has the promise of eventually providing an interpretable, low-dimensional personality representation, which is also highly predictive of behavior.
| S-EPMC9314364 | biostudies-literature
Project description:MotivationProtein engineering techniques are key in designing novel catalysts for a wide range of reactions. Although approaches vary in their exploration of the sequence-structure-function paradigm, they are often hampered by the labor-intensive steps of protein expression and screening. In this work, we describe the development and testing of a high-throughput in silico sequence-structure-function pipeline using AlphaFold2 and fast Fourier transform docking that is benchmarked with enantioselectivity and reactivity predictions for an ancestral sequence library of fungal flavin-dependent monooxygenases.ResultsThe predicted enantioselectivities and reactivities correlate well with previously described screens of an experimentally available subset of these proteins and capture known changes in enantioselectivity across the phylogenetic tree representing ancestorial proteins from this family. With this pipeline established as our functional screen, we apply ensemble decision tree models and explainable AI techniques to build sequence-function models and extract critical residues within the binding site and the second-sphere residues around this site. We demonstrate that the top-identified key residues in the control of enantioselectivity and reactivity correspond to experimentally verified residues. The in silico sequence-to-function pipeline serves as an accelerated framework to inform protein engineering efforts from vast informative sequence landscapes contained in protein families, ancestral resurrects, and directed evolution campaigns.AvailabilityJupyter notebooks detailing the sequence-structure-function pipeline are available at https://github.com/BrooksResearchGroup-UM/seq_struct_func.
| S-EPMC10789314 | biostudies-literature
Project description:The ability to assign function to proteins has become a major bottleneck for comprehensively understanding cellular mechanisms at the molecular level. Here we discuss the extent to which structural domain classifications can help in deciphering the complex relationship between the functions of proteins and their sequences and structures. Structural classifications are particularly helpful in understanding the mosaic manner in which new proteins and functions emerge through evolution. This is partly because they provide reliable and concrete domain definitions and enable the detection of very remote structural similarities and homologies. It is also because structural data can illuminate more clearly the mechanisms by which a broader functional repertoire can emerge during evolution.
| S-EPMC2920418 | biostudies-literature
Project description:Osteosarcoma is the most common bone sarcoma in children and young adults. While chemotherapy is universally delivered, benefit from chemotherapy is limited to roughly half of localized patients. Increasingly, intratumoral heterogeneity is being appreciated as a source of therapeutic resistance. In this study we developed and evaluated an in vitro model of osteosarcoma heterogeneity, characterizing phenotype (growth in varying environments, sensitivity to chemotherapy) and genotype. We present the genotypic and phenotypic characterization of an osteosarcoma cell line panel with a focus on coculture of the most phenotypically divergent cell lines, 143B and SAOS2. The extent of phenotypic heterogeneity can be altered with relatively modest environmental (pH, glutamine) or chemical perturbations. We demonstrate that in nutrient rich in vitro culture conditions 143B outcompetes SAOS2, but with selection pressure from nutrient variations or conventional chemotherapy, SAOS2 growth can be favored in spheroids. Importantly, perturbations that affect the faster growing cell line have only a modest effect on final spheroid size when the simplest heterogeneity state (a two-cell line coculture) is evaluated, and thus the only evaluated therapies to eliminate the spheroids were by switching therapies from a first strike to a second strike. This extensively characterized, widely available system can be modeled and scaled to allow for improved strategies to anticipate resistance in osteosarcoma due to phenotypic heterogeneity.
2023-11-22 | GSE240278 | GEO
Project description:Predicting the functional roles of proteins based on various genome-wide data, such as protein-protein association networks, has become a canonical problem in computational biology. Approaching this task as a binary classification problem, we develop a network-based extension of the spatial auto-probit model. In particular, we develop a hierarchical Bayesian probit-based framework for modeling binary network-indexed processes, with a latent multivariate conditional autoregressive Gaussian process. The latter allows for the easy incorporation of protein-protein association network topologies-either binary or weighted-in modeling protein functional similarity. We use this framework to predict protein functions, for functions defined as terms in the Gene Ontology (GO) database, a popular rigorous vocabulary for biological functionality. Furthermore, we show how a natural extension of this framework can be used to model and correct for the high percentage of false negative labels in training data derived from GO, a serious shortcoming endemic to biological databases of this type. Our method performance is evaluated and compared with standard algorithms on weighted yeast protein-protein association networks, extracted from a recently developed integrative database called Search Tool for the Retrieval of INteracting Genes/proteins (STRING). Results show that our basic method is competitive with these other methods, and that the extended method-incorporating the uncertainty in negative labels among the training data-can yield nontrivial improvements in predictive accuracy.
| S-EPMC3116961 | biostudies-literature