Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Protein language model”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 127 records · Page 7Linked to original sources
PubMed2002

AstexViewer: a visualisation aid for structure-based drug design.

AstexViewer is a Java molecular graphics program that can be used for visualisation in many aspects of structure-based drug design. This paper describes its functionality, implementation and examples of its use. The program can run as an Applet in a web browser allowing structures to be displayed without installing additional software. Applications of its use are described for visualisation and as part of a structure based design platform. The software is being made freely available to the community and may be downloaded from http://www.astex-technology.com/AstexViewer.

Computer Graphics↗
PubMed2005

A comprehensive pathway map of epidermal growth factor receptor signaling.

The epidermal growth factor receptor (EGFR) signaling pathway is one of the most important pathways that regulate growth, survival, proliferation, and differentiation in mammalian cells. Reflecting this importance, it is one of the best-investigated signaling systems, both experimentally and computationally, and several computational models have been developed for dynamic analysis. A map of molecular interactions of the EGFR signaling system is a valuable resource for research in this area. In this paper, we present a comprehensive pathway map of EGFR signaling and other related pathways. The map reveals that the overall architecture of the pathway is a bow-tie (or hourglass) structure with several feedback loops. The map is created using CellDesigner software that enables us to graphically represent interactions using a well-defined and consistent graphical notation, and to store it in Systems Biology Markup Language (SBML).

Animals↗
PubMed2006

Statistical modeling of biomedical corpora: mining the Caenorhabditis Genetic Center Bibliography for genes related to life span.

BACKGROUND: The statistical modeling of biomedical corpora could yield integrated, coarse-to-fine views of biological phenomena that complement discoveries made from analysis of molecular sequence and profiling data. Here, the potential of such modeling is demonstrated by examining the 5,225 free-text items in the Caenorhabditis Genetic Center (CGC) Bibliography using techniques from statistical information retrieval. Items in the CGC biomedical text corpus were modeled using the Latent Dirichlet Allocation (LDA) model. LDA is a hierarchical Bayesian model which represents a document as a random mixture over latent topics; each topic is characterized by a distribution over words. RESULTS: An LDA model estimated from CGC items had better predictive performance than two standard models (unigram and mixture of unigrams) trained using the same data. To illustrate the practical utility of LDA models of biomedical corpora, a trained CGC LDA model was used for a retrospective study of nematode genes known to be associated with life span modification. Corpus-, document-, and word-level LDA parameters were combined with terms from the Gene Ontology to enhance the explanatory value of the CGC LDA model, and to suggest additional candidates for age-related genes. A novel, pairwise document similarity measure based on the posterior distribution on the topic simplex was formulated and used to search the CGC database for "homologs" of a "query" document discussing the life span-modifying clk-2 gene. Inspection of these document homologs enabled and facilitated the production of hypotheses about the function and role of clk-2. CONCLUSION: Like other graphical models for genetic, genomic and other types of biological data, LDA provides a method for extracting unanticipated insights and generating predictions amenable to subsequent experimental validation.

Animals↗
PubMed2005

Tools enabling the elucidation of molecular pathways active in human disease: application to Hepatitis C virus infection.

BACKGROUND: The extraction of biological knowledge from genome-scale data sets requires its analysis in the context of additional biological information. The importance of integrating experimental data sets with molecular interaction networks has been recognized and applied to the study of model organisms, but its systematic application to the study of human disease has lagged behind due to the lack of tools for performing such integration. RESULTS: We have developed techniques and software tools for simplifying and streamlining the process of integration of diverse experimental data types in molecular networks, as well as for the analysis of these networks. We applied these techniques to extract, from genomic expression data from Hepatitis C virus-infected liver tissue, potentially useful hypotheses related to the onset of this disease. Our integration of the expression data with large-scale molecular interaction networks and subsequent analyses identified molecular pathways that appear to be induced or repressed in the response to Hepatitis C viral infection. CONCLUSION: The methods and tools we have implemented allow for the efficient dynamic integration and analysis of diverse data in a major human disease system. This integrated data set in turn enabled simple analyses to yield hypotheses related to the response to Hepatitis C viral infection.

Computational Biology↗
PubMed2005

SplitTester: software to identify domains responsible for functional divergence in protein family.

BACKGROUND: Many protein families have undergone functional divergence after gene duplications such that current subgroups of the family carry out overlapping but distinct biological roles. For the protein families with known functional subtypes (a functional split), we developed the software, SplitTester, to identify potential regions that are responsible for the observed distinct functional subtypes within the same protein family. RESULTS: Our software, SplitTester, takes a multiple protein sequences alignment as input, generated from protein members of two subgroups with known functional divergence. SplitTester was designed to construct the neighbor joining tree (a split cluster) from variable-sized sliding windows across the alignment in a process called split-clustering. SplitTester identifies the regions, whose split cluster is consistent with the functional split, but may be inconsistent with the phylogeny of the protein family. We hypothesize that at least some number of these identified regions, which are not following a random mutation process, are responsible for the observed functional split. To test our method, we used reverse transcriptase from a group of Pseudoviridae retrotransposons: to identify residues specific for diverged primer recognition. Candidate regions were then mapped onto the three dimensional structures of reverse transcriptase. The locations of these amino acids within the enzyme are consistent with their biological roles. CONCLUSION: SplitTester aims to identify specific domain sequences responsible for functional divergence of subgroups within a protein family. From the analysis of retroelements reverse transcriptase family, we successfully identified the regions splitting this family according to the primer specificity, implying their functions in the specific primer selection.

Algorithms↗
PubMed2006

Joint learning of gene functions--a Bayesian network model approach.

In this paper, we develop a machine learning system for determining gene functions from heterogeneous data sources using a Weighted Naive Bayesian network (WNB). The knowledge of gene functions is crucial for understanding many fundamental biological mechanisms such as regulatory pathways, cell cycles and diseases. Our major goal is to accurately infer functions of putative genes or Open Reading Frames (ORFs) from existing databases using computational methods. However, this task is intrinsically difficult since the underlying biological processes represent complex interactions of multiple entities. Therefore, many functional links would be missing when only one or two sources of data are used in the prediction. Our hypothesis is that integrating evidence from multiple and complementary sources could significantly improve the prediction accuracy. In this paper, our experimental results not only suggest that the above hypothesis is valid, but also provide guidelines for using the WNB system for data collection, training and predictions. The combined training data sets contain information from gene annotations, gene expressions, clustering outputs, keyword annotations, and sequence homology from public databases. The current system is trained and tested on the genes of budding yeast Saccharomyces cerevisiae. Our WNB model can also be used to analyze the contribution of each source of information toward the prediction performance through the weight training process. The contribution analysis could potentially lead to significant scientific discovery by facilitating the interpretation and understanding of the complex relationships between biological entities.

Artificial Intelligence↗
PubMed2005

Linking enzyme sequence to function using Conserved Property Difference Locator to identify and annotate positions likely to control specific functionality.

BACKGROUND: Families of homologous enzymes evolved from common progenitors. The availability of multiple sequences representing each activity presents an opportunity for extracting information specifying the functionality of individual homologs. We present a straightforward method for the identification of residues likely to determine class specific functionality in which multiple sequence alignments are converted to an annotated graphical form by the Conserved Property Difference Locator (CPDL) program. RESULTS: Three test cases, each comprised of two groups of functionally-distinct homologs, are presented. Of the test cases, one is a membrane and two are soluble enzyme families. The desaturase/hydroxylase data was used to design and test the CPDL algorithm because a comparative sequence approach had been successfully applied to manipulate the specificity of these enzymes. The other two cases, ATP/GTP cyclases, and MurD/MurE synthases were chosen because they are well characterized structurally and biochemically. For the desaturase/hydroxylase enzymes, the ATP/GTP cyclases and the MurD/MurE synthases, groups of 8 (of approximately 400), 4 (of approximately 150) and 10 (of >400) residues, respectively, of interest were identified that contain empirically defined specificity determining positions. CONCLUSION: CPDL consistently identifies positions near enzyme active sites that include those predicted from structural and/or biochemical studies to be important for specificity and/or function. This suggests that CPDL will have broad utility for the identification of potential class determining residues based on multiple sequence analysis of groups of homologous proteins. Because the method is sequence, rather than structure, based it is equally well suited for designing structure-function experiments to investigate membrane and soluble proteins.

Algorithms↗
PubMed1998

From sequence to structure to literature: the protocol approach to bioinformation.

Until the recent advent of high-throughput experimental data-acquisition in biology, the computational analysis of the biological data was predominantly on an ad hoc basis--i.e., the application of a given piece of software on the biological data depended on the need of the moment. This "functional approach" often resulted in piecemeal computational analysis with large amount of intervening "dead-time". The present high-throughput availability of experimental biological data requires a more streamlined and integrated "protocol approach". In this work, we illustrate such a user-friendly protocol using a common question frequently faced by a wet-lab bench-biologist--"Now that I have a DNA or protein sequence, what can I do with it using a computer?" As phrased, this question is steeped in the functional approach. In contrast, the protocol approach would re-phrase the same question as "Now that I have a DNA or protein sequence, what can a computer do for me?" Our integrating tool can start with a sequence and build a substantial custom data-warehouse of computationally derived sequence information, structure information and relevant published literature, that is continually updated.

Amino Acid Sequence↗
PubMed2026

Artificial Intelligence for Natural Products Discovery and Development.

Natural products (NPs) remain a cornerstone of modern drug discovery, offering stereochemical complexity and diverse bioactivities that precisely modulate therapeutic targets, refined through billions of years of evolution. However, their research has long been hindered by inefficient, empirical workflows, high resource consumption, structural complexity, and the "multicomponent, multi-target" nature of their mechanisms. The exponential growth of genomic, metabolomic, and spectral data has overwhelmed conventional analytical methods, exposing critical bottlenecks in handling high-dimensional, heterogeneous datasets that exceed human interpretive capacity. Artificial intelligence (AI) is emerging as a transformative paradigm to address these challenges, integrating multi-omics and chemical data to shift NP research from fragmented empiricism toward mechanism-driven, precision-oriented development. By leveraging deep learning architectures- including graph neural networks, Transformers, and diffusion-based generative models-AI enables systematic decoding of NP biosynthesis, automated structure elucidation, rational target identification, knowledge extraction from vast unstructured scientific literature, and de novo molecular design. This review comprehensively surveys recent advances in AI applications across the full NP discovery and development pipeline, encompassing genome mining, structure-based and ligand-based virtual screening, multimodal structural characterization, lead optimization, and biosynthetic pathway engineering. We further examine the emerging roles of protein-centric, molecule- centric, and multimodal foundation models, as well as large language models, in bridging genotype-to-chemotype gaps and unlocking unstructured scientific knowledge. Finally, we discuss critical challenges including data scarcity, representational limitations for complex stereochemistry, physical plausibility in generative models, and the urgent need for experimental validation, while outlining future directions toward autonomous experimentation, closed-loop optimization, and human-AI collaborative discovery.

Artificial intelligence↗
PubMed2006

Creation and implications of a phenome-genome network.

Although gene and protein measurements are increasing in quantity and comprehensiveness, they do not characterize a sample's entire phenotype in an environmental or experimental context. Here we comprehensively consider associations between components of phenotype, genotype and environment to identify genes that may govern phenotype and responses to the environment. Context from the annotations of gene expression data sets in the Gene Expression Omnibus is represented using the Unified Medical Language System, a compendium of biomedical vocabularies with nearly 1-million concepts. After showing how data sets can be clustered by annotative concepts, we find a network of relations between phenotypic, disease, environmental and experimental contexts as well as genes with differential expression associated with these concepts. We identify novel genes related to concepts such as aging. Comprehensively identifying genes related to phenotype and environment is a step toward the Human Phenome Project.

Aging↗
PubMed2005

Discovery of protein-protein interactions using a combination of linguistic, statistical and graphical information.

BACKGROUND: The rapid publication of important research in the biomedical literature makes it increasingly difficult for researchers to keep current with significant work in their area of interest. RESULTS: This paper reports a scalable method for the discovery of protein-protein interactions in Medline abstracts, using a combination of text analytics, statistical and graphical analysis, and a set of easily implemented rules. Applying these techniques to 12,300 abstracts, a precision of 0.61 and a recall of 0.97 were obtained, (f = 0.74) and when allowing for two-hop and three-hop relations discovered by graphical analysis, the precision was 0.74 (f = 0.83). CONCLUSION: This combination of linguistic and statistical approaches appears to provide the highest precision and recall thus far reported in detecting protein-protein relations using text analytic approaches.

Artificial Intelligence↗
PubMedDate not supplied

Genomic messaging system and DNA mark-up language for information-based personalized medicine with clinical and proteome research applications.

The convergence of clinical medicine and the Life Sciences, commencing with opportunities in clinical trials and clinically linked medical research, presents many novel challenges. The Genomic Messaging System (GMS) described here was originally developed as a tool for assembling clinical genomic records of individual and collective patients, and was then generalized to become a flexible workflow component that will link clinical records to a variety of computational biology research tools, for research and ultimately for a more personalized, focused, and preventative healthcare system. Prominent among the applications linked are protein science applications, including the rapid automated modeling of patient proteins with their individual structural polymorphisms. In an initial study, GMS formed the basis of a fully automated system for modeling patient proteins with structural polymorphisms as a basis for drug selection and ultimately design on an individual patient basis.

Clinical Medicine↗
PubMed2006

Combining evidence, biomedical literature and statistical dependence: new insights for functional annotation of gene sets.

BACKGROUND: Large-scale genomic studies based on transcriptome technologies provide clusters of genes that need to be functionally annotated. The Gene Ontology (GO) implements a controlled vocabulary organised into three hierarchies: cellular components, molecular functions and biological processes. This terminology allows a coherent and consistent description of the knowledge about gene functions. The GO terms related to genes come primarily from semi-automatic annotations made by trained biologists (annotation based on evidence) or text-mining of the published scientific literature (literature profiling). RESULTS: We report an original functional annotation method based on a combination of evidence and literature that overcomes the weaknesses and the limitations of each approach. It relies on the Gene Ontology Annotation database (GOA Human) and the PubGene biomedical literature index. We support these annotations with statistically associated GO terms and retrieve associative relations across the three GO hierarchies to emphasise the major pathways involved by a gene cluster. Both annotation methods and associative relations were quantitatively evaluated with a reference set of 7397 genes and a multi-cluster study of 14 clusters. We also validated the biological appropriateness of our hybrid method with the annotation of a single gene (cdc2) and that of a down-regulated cluster of 37 genes identified by a transcriptome study of an in vitro enterocyte differentiation model (CaCo-2 cells). CONCLUSION: The combination of both approaches is more informative than either separate approach: literature mining can enrich an annotation based only on evidence. Text-mining of the literature can also find valuable associated MEDLINE references that confirm the relevance of the annotation. Eventually, GO terms networks can be built with associative relations in order to highlight cooperative and competitive pathways and their connected molecular functions.

Algorithms↗
PubMed2006

Optimized mixed Markov models for motif identification.

BACKGROUND: Identifying functional elements, such as transcriptional factor binding sites, is a fundamental step in reconstructing gene regulatory networks and remains a challenging issue, largely due to limited availability of training samples. RESULTS: We introduce a novel and flexible model, the Optimized Mixture Markov model (OMiMa), and related methods to allow adjustment of model complexity for different motifs. In comparison with other leading methods, OMiMa can incorporate more than the NNSplice's pairwise dependencies; OMiMa avoids model over-fitting better than the Permuted Variable Length Markov Model (PVLMM); and OMiMa requires smaller training samples than the Maximum Entropy Model (MEM). Testing on both simulated and actual data (regulatory cis-elements and splice sites), we found OMiMa's performance superior to the other leading methods in terms of prediction accuracy, required size of training data or computational time. Our OMiMa system, to our knowledge, is the only motif finding tool that incorporates automatic selection of the best model. OMiMa is freely available at 1. CONCLUSION: Our optimized mixture of Markov models represents an alternative to the existing methods for modeling dependent structures within a biological motif. Our model is conceptually simple and effective, and can improve prediction accuracy and/or computational speed over other leading methods.

Algorithms↗
PubMed1997

Using Tcl for molecular visualization and analysis.

Reading and manipulating molecular structure data is a standard task in every molecular visualization and analysis program, but is rarely available in a form readily accessible to the user. Instead, the development of new methods for analysis, display, and interaction is often achieved by writing a new program, rather than building on pre-existing software. We present the Tcl-based script language used in our molecular modeling program, VMD, and show how it can access information about the molecular structure, perform analysis, and graphically display and animate the results. The commands are available to the user and make VMD a useful environment for studying biomolecules.

Computer Simulation↗
PubMed2026

Deep generative models in biological sequence and structure analysis and design.

Deep generative models have transformed biological sequence modeling from predictive analysis toward increasingly controllable design. Early biological applications of Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs) established latent representation learning and sequence synthesis, while recent advances in transformer-based language models, discrete diffusion, flow-matching, and multimodal generative frameworks have substantially expanded the scope of biological design. This review examines generative models for DNA, RNA, and protein sequence design, emphasizing how different model classes represent biological constraints, operate over discrete and continuous spaces, and integrate sequence, structure, and function. We compare VAEs, GANs, autoregressive and masked language models, diffusion models, and flow-based approaches across genomics, transcriptomics, and proteomics, with particular attention to controllability, long-range dependency modeling, structural grounding, generalization, and experimental utility. We further examine evaluation strategies, out-of-distribution generalization, and closed-loop design-build-test-learn workflows that connect in silico generation with empirical validation. We distinguish fundamental modality-dependent constraints including sequence discreteness, context length, structural coupling, and physical or thermodynamic requirements from architecture-dependent advantages that reflect the current state of the field. Current studies suggest that long-context models are particularly useful for genome-scale representation and sequence modeling, whereas structure-aware diffusion, flow-based, and inverse-folding approaches provide better frameworks for geometry-constrained RNA and protein design. This perspective provides a critical framework for understanding the present capabilities, limitations, and convergence of generative approaches toward reliable and experimentally grounded biological design.

Biological sequence analysis↗
PubMed2004

Delineation and analysis of the conceptual data model implied by the "IUPAC Recommendations for Biochemical Nomenclature".

Computational analysis of the bonding, geometric, and topological relationships within proteins typically takes on the order of hours, mainly devoted to the writing of scripts and code to correctly parse the data. The Structured Query Language (SQL) built into modern database management systems eliminates the need for data parsing, effectively reducing the analysis time to seconds. To this end, we have formulated a conceptual data model (CDM) for proteins based on the IUPAC recommendations for biochemical nomenclature. This conceptual data model makes explicit the inherent bonding relationships between the atoms of a protein, as well as the geometric (bond angle and torsion angle) and topological (chirality) relationships between the bonds. The validity of the CDM has been tested with a reduced implementation using commercial database software. The ease in both populating the database with data from the Protein Data Bank and formulating/executing queries supports the correctness of the model. The ability to conduct truly interactive analyses of protein structure is essential to fully capitalize on the explosion in postgenomic protein structure data.

Biochemical Phenomena↗
PubMed2004

Superfamilies of evolved and designed networks.

Complex biological, technological, and sociological networks can be of very different sizes and connectivities, making it difficult to compare their structures. Here we present an approach to systematically study similarity in the local structure of networks, based on the significance profile (SP) of small subgraphs in the network compared to randomized networks. We find several superfamilies of previously unrelated networks with very similar SPs. One superfamily, including transcription networks of microorganisms, represents "rate-limited" information-processing networks strongly constrained by the response time of their components. A distinct superfamily includes protein signaling, developmental genetic networks, and neuronal wiring. Additional superfamilies include power grids, protein-structure networks and geometric networks, World Wide Web links and social networks, and word-adjacency networks from different languages.

Animals↗