PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “genomic language model”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 73 records · Page 4Linked to original sources
PubMed2005

Automatic detection of subsystem/pathway variants in genome analysis.

MOTIVATION: Proteins work together in pathways and networks, collectively comprising the cellular machinery. A subsystem (a generalization of pathway concept) is a group of related functional roles (such as enzymes) jointly involved in a specific aspect of the cellular machinery. Subsystems provide a natural framework for comparative genome analysis and functional annotation. A subsystem may be implemented in a number of different functional variants in individual species. In order to reliably project functional assignments across multiple genomes, we have to be able to identify the variants implemented in each genome. The analysis of such variants across diverse species is an interesting problem by itself and may provide new evolutionary insights. However, no computational techniques are presently available for an automated detection and analysis of subsystem variants. RESULTS: Here we formulate the subsystem variant detection problem as finding the minimum number of subgraphs of a subsystem, which is represented as a graph, and solve the optimization problem by integer programming approach. The performance of our method was tested on subsystems encoded in the SEED, a genomic integration platform developed by the Fellowship for Interpretation of Genomes as a component of a large-scale effort on comparative analysis and annotation of multiple diverse genomes. Here we illustrate the results obtained for two expert-encoded subsystems of the biosynthesis of Coenzyme A and FMN/FAD cofactors. Applications of variant detection, to support genomic annotations and to assess divergence of species, are briefly discussed in the context of these universally conserved and essential metabolic subsystems. SUPPLEMENTARY INFORMATION: The details of the variant detection results are available at http://ffas.burnham.org/svar/supp.html.

Animals↗
PubMed2006

SNAP: Combine and Map modules for multilocus population genetic analysis.

We have added two software tools to our Suite of Nucleotide Analysis Programs (SNAP) for working with DNA sequences sampled from populations. SNAP Map collapses DNA sequence data into unique haplotypes, extracts variable sites and manipulates output into multiple formats for input into existing software packages for evolutionary analyses. Map collapses DNA sequence data into unique haplotypes, extracts variable sites and manipulates output into multiple formats for input into existing software packages for evolutionary analyses. Map includes novel features such as recoding insertions or deletions, including or excluding variable sites that violate an infinite-sites model and the option of collapsing sequences with corresponding phenotypic information, important in testing for significant haplotype-phenotype associations. SNAP Combine merges multiple DNA sequence alignments into a single multiple alignment file. The resulting file can be the union or intersection of the input files. SNAP Combine currently reads from and writes to several sequence alignment file formats including both sequential and interleaved formats. Combine also keeps track of the start and end positions of each separate alignment file allowing the user to exclude variable sites or taxa, important in creating input files for multilocus analyses.

Algorithms↗
PubMed2026

Rational and computation-assisted engineering of a compact and efficient CRISPR-Cas12f genome editor.

The CRISPR-Cas12f system is an ultracompact genome-editing platform, yet only a few orthologs exhibit robust activity in mammalian cells. Here, we systematically screened 23 Cas12f orthologs and identified two active nucleases, PspCas12f1 and TcCas12f1, capable of genome editing in human cells. Single guide RNA (sgRNA) scaffold optimization enhanced the basal activity of PspCas12f1. To further improve its performance, we combined structure-guided rational design with protein language model-assisted filtering. Candidate mutations predicted by SaProt were further screened based on structural proximity to the DNA-binding interface and electrostatic compatibility. This integrative strategy identified Q100R and E293R, whose combination yielded the optimized variant enPspCas12f1. enPspCas12f1 achieved genome-editing efficiencies comparable to SpCas9 across multiple endogenous loci while maintaining high specificity. Collectively, our results demonstrate that integrating protein language model-assisted filtering with structure-guided rational design provides an effective strategy for engineering PspCas12f1 and may facilitate the optimization of additional compact CRISPR nucleases.

CRISPR-Cas12f↗
PubMed2025

Embed-Search-Align: DNA sequence alignment using Transformer models.

MOTIVATION: DNA sequence alignment, an important genomic task, involves assigning short DNA reads to the most probable locations on an extensive reference genome. Conventional methods tackle this challenge in two steps: genome indexing followed by efficient search to locate likely positions for given reads. Building on the success of Large Language Models in encoding text into embeddings, where the distance metric captures semantic similarity, recent efforts have encoded DNA sequences into vectors using Transformers and have shown promising results in tasks involving classification of short DNA sequences. Performance at sequence classification tasks does not, however, guarantee sequence alignment, where it is necessary to conduct a genome-wide search to align every read successfully, a significantly longer-range task by comparison. RESULTS: We bridge this gap by developing a "Embed-Search-Align" (ESA) framework, where a novel Reference-Free DNA Embedding (RDE) Transformer model generates vector embeddings of reads and fragments of the reference in a shared vector space; read-fragment distance metric is then used as a surrogate for sequence similarity. ESA introduces: (i) Contrastive loss for self-supervised training of DNA sequence representations, facilitating rich reference-free, sequence-level embeddings, and (ii) a DNA vector store to enable search across fragments on a global scale. RDE is 99% accurate when aligning 250-length reads onto a human reference genome of 3 gigabases (single-haploid), rivaling conventional algorithmic sequence alignment methods such as Bowtie and BWA-Mem. RDE far exceeds the performance of six recent DNA-Transformer model baselines such as Nucleotide Transformer, Hyena-DNA, and shows task transfer across chromosomes and species. AVAILABILITY AND IMPLEMENTATION: Please see https://anonymous.4open.science/r/dna2vec-7E4E/readme.md.

Sequence Analysis, DNA↗
PubMed2002

Impact of genomics approaches on plant genetics and physiology.

Comprehensive analysis of genetic information in higher plants is under way for several plants of biological and agronomical importance. Among them, Arabidopsis thaliana, a member of Brassica family, and Oryza sativa(rice) have been chosen as model plants most suitable for genome analysis. Sequencing of the genome of A. thaliana was completed in December 2000, and rice genome sequencing is in progress. The accumulated genome sequences, together with the hundreds of thousands of ESTs from several tens of plant species, have drastically changed the strategy of plant genetics. By utilizing the information on the genome and gene structures, comprehensive approaches for genome-wide functional analysis of the genes, including transcriptome analysis using microarray systems and a comprehensive analysis of a large number of insertion mutant lines, have been widely adopted. As a consequence, a large quantity of information on both the structure and function of genes in these model plants has been accumulated. However, other plant species may have their own characteristics and advantages to study individual phenomena. Application of knowledge from the model plants to other plant species and vice versa through the common language, namely the genome information, should facilitate understanding of the genetic systems underlying a variety of biological phenomena. Introduction of this common language may not be very simple, especially in the case of complex pathways such as a process of cell-covering formation. Nevertheless, it should be emphasized that genomics approaches are the most promising way to understand these processes.

Journal Article↗
PubMed2000

Genetics of kidney disease.

Perhaps nothing in the fields of medicine and nephrology is moving more rapidly than genetics. From this movement are opportunities for discovery, new therapy, and better counseling for patients. At a level of basic science, renal medicine has been a consistent contributor to this emerging discipline, but our current approach to training in the methods and uses of human genetics probably will not keep up with the technology, nor the needs of the modern bedside practitioner. The facile use of genetics in the next century will require the construction and exploration of new disease models, rededication to human informatics, and teaching the language of molecular and population genomics.

Erythropoietin↗
PubMed2001

Simultaneous modelling of metabolic, genetic and product-interaction networks.

The creation of cell models from annotated genome information, as well as additional data from other databases, requires both a format and medium for its distribution. Standards are described for the representation of the data in the form of Document Type Definitions (DTDs) for XML files. Separate DTDs are detailed for genetic, metabolic and gene product-interaction networks, which can be used to hold information on individual subsystems, or which may be combined to create a whole cell DTD. In the execution of this work, a fifth DTD was also created for a metabolite thesaurus, which allows incorporation of metabolite synonyms and generic nomenclature data into the models. A gene-regulation classification scheme was also created, to facilitate incorporation of gene regulatory information in an efficient manner. The work is described with particular reference to the metabolic network of Escherichia coli, which contains 808 individual enzymes. The assignment of confidence levels to these data, through the use of Gene Ontology evidence codes, is highlighted. In silico investigations may now be performed using the mathematical simulation workbench, DBsolve, which incorporates the facility to introduce data directly from XML.

Computational Biology↗
PubMed2026

Agentic AI for Spatial Omics.

This highlight summarises recent advances in agentic artificial intelligence (AI) systems for spatial omics analysis. These systems are compared along two central tensions: autonomy versus accountability, and adaptability versus reproducibility. We argue that progress will depend not on maximising automation, but on defining where autonomy is appropriate.

Artificial Intelligence↗
PubMed2005

Evolutionary sequence analysis of complete eukaryote genomes.

BACKGROUND: Gene duplication and gene loss during the evolution of eukaryotes have hindered attempts to estimate phylogenies and divergence times of species. Although current methods that identify clusters of orthologous genes in complete genomes have helped to investigate gene function and gene content, they have not been optimized for evolutionary sequence analyses requiring strict orthology and complete gene matrices. Here we adopt a relatively simple and fast genome comparison approach designed to assemble orthologs for evolutionary analysis. Our approach identifies single-copy genes representing only species divergences (panorthologs) in order to minimize potential errors caused by gene duplication. We apply this approach to complete sets of proteins from published eukaryote genomes specifically for phylogeny and time estimation. RESULTS: Despite the conservative criterion used, 753 panorthologs (proteins) were identified for evolutionary analysis with four genomes, resulting in a single alignment of 287,000 amino acids. With this data set, we estimate that the divergence between deuterostomes and arthropods took place in the Precambrian, approximately 400 million years before the first appearance of animals in the fossil record. Additional analyses were performed with seven, 12, and 15 eukaryote genomes resulting in similar divergence time estimates and phylogenies. CONCLUSION: Our results with available eukaryote genomes agree with previous results using conventional methods of sequence data assembly from genomes. They show that large sequence data sets can be generated relatively quickly and efficiently for evolutionary analyses of complete genomes.

Animals↗
PubMed1999

An automated comparative analysis of 17 complete microbial genomes.

MOTIVATION: As sequenced genomes become larger and sequencing becomes faster, there is a need to develop accurate automated genome comparison techniques and databases to facilitate derivation of genome functionality; identification of enzymes, putative operons and metabolic pathways; and to derive phylogenetic classification of microbes. RESULTS: This paper extends an automated pair-wise genome comparison technique (Bansal et al., Math. Model. Sci. Comput., 9, 1-23, 1998, Bansal and Bork, in First International Workshop of Declarative Languages, Springer, pp. 275-289, 1999) used to identify orthologs and gene groups to derive orthologous genes in a group of genomes and to identify genes with conserved functionality. Seventeen microbial genomes archived at ftp://ncbi.nlm.nih.gov/genbank/genomes have been compared using the automated technique. Data related to orthologs, gene groups, gene duplication, gene fusion, orthologs with conserved functionality, and genes specifically orthologous to Escherichia coli and pathogens has been presented and analyzed. AVAILABILITY: A prototype database is available at ftp://www.mcs.kent.edu/arvind/intellibio / orthos.html. The software is free for academic research under an academic license. The detailed database for every microbial genome in NCBI is commercially available through intellibio software and consultancy corporation (Web site: http://www.mcs.kent.edu/årvind/intellibio . html). CONTACT: arvind@mcs.kent.edu.

Algorithms↗
PubMed2025

Pre-Meta: priors-augmented retrieval for LLM-based metadata generation.

MOTIVATION: While high-throughput sequencing technologies have dramatically accelerated genomic data generation, the manual processes required for dataset annotation and metadata creation impede the efficient discovery and publication of these resources across disparate public repositories. Large language models (LLMs) have the potential to streamline dataset profiling and discovery. However, their current limitations in generalizing across specialized knowledge domains, particularly in fields such as biomedical genomics, prevent them from fully realizing this potential. This article presents Pre-Meta, an LLM-agnostic and domain-independent data annotation pipeline with an enriched retrieval procedure that leverages related priors-such as pre-generated metadata tags and ontologies-as auxiliary information to improve the accuracy of automated metadata generation. RESULTS: Validated using five selected metadata fields sampled across 1500 papers, the Pre-Meta assisted annotation experiment-without finetuning and prompt optimization-demonstrates a systemic improvement in the annotation task: shown through a 23%, 72%, and 75% accuracy gain from conventional RAG adoptions of GPT-4o mini, Llama 8B, and Mistral 7B respectively. AVAILABILITY AND IMPLEMENTATION: The code, data access, and scripts are available at: https://github.com/SINTEF-SE/LLMDap.

Metadata↗
PubMed2026

Accelerating inference in genomic and proteomic foundation models via speculative decoding.

MOTIVATION: Genomic and protein foundation models (GFMs and PFMs) have demonstrated strong performance in learning the language of DNA and proteins, but their use in large-scale sequence generation is limited by the latency of autoregressive decoding. Because every token triggers a forward pass of a large Transformer, whose inference is relatively slow, long-sequence generation quickly becomes costly. RESULTS: In this work we adapt speculative decoding to a representative GFM: the DNA model DNAGPT and two representative PFMs: ProGen2 and ProtGPT2. We implement a probabilistic variant of speculative decoding, in which a lightweight draft model proposes short token spans and a larger target model verifies or corrects them in parallel, while preserving the target model's sampling distribution. Across all three models we systematically study the effect of speculation window length, temperature, draft architecture and prompt length, and we benchmark tokens per second over multiple runs per configuration. Speculative decoding yields consistent speedups over standard key-value cached decoding, with maximum observed speedup reaching 100% increase, while average gains across models ranging between 20% and 40% (e.g. 1.2×-1.4×), without changing the underlying target model predictions. Our results show that speculative decoding is a practical and model-agnostic strategy for accelerating genomic and proteomic sequence generation without sacrificing prediction quality. AVAILABILITY AND IMPLEMENTATION: All code and results are freely available at https://github.com/Georgakopoulos-Soares-lab/BioSpecDec.

Genomics↗
PubMed2026

AI-driven CRISPR screening: optimizing gene editing through automation and intelligent decision support.

BACKGROUND: CRISPR-based genetic screening has become a central methodology in functional genomics, enabling systematic interrogation of gene function, genetic interactions and context-dependent vulnerabilities at scale. However, the rapid expansion of screening modalities-including multi-condition designs, combinatorial perturbations, in vivo applications and single-cell readouts-has exposed fundamental limitations of heuristic-driven experimental design and post hoc statistical analysis. MAIN BODY: This Review synthesizes how artificial intelligence is reshaping CRISPR screening by introducing predictive, adaptive and system-level intelligence across the experimental lifecycle. We organize recent advances into two tightly coupled modules. First, machine learning and deep learning (ML/DL) methods optimize experimental design by learning context-dependent perturbation behavior, anticipating confounding effects and enabling iterative, information-efficient screening strategies. Second, large language model-agent (LLM-agent) systems complement these advances by externalizing scientific reasoning, integrating biological knowledge at scale and coordinating analysis and decision-making in human-in-the-loop workflows. CONCLUSIONS: Together, ML/DL and LLM-agent approaches reframe CRISPR screening from a static analytical pipeline into an intelligent experimental system, with important implications for robustness, scalability and biological discovery.

Artificial Intelligence↗
PubMed2006

The eloquent ape: genes, brains and the evolution of language.

The human capacity to acquire complex language seems to be without parallel in the natural world. The origins of this remarkable trait have long resisted adequate explanation, but advances in fields that range from molecular genetics to cognitive neuroscience offer new promise. Here we synthesize recent developments in linguistics, psychology and neuroimaging with progress in comparative genomics, gene-expression profiling and studies of developmental disorders. We argue that language should be viewed not as a wholesale innovation, but as a complex reconfiguration of ancestral systems that have been adapted in evolutionarily novel ways.

Animal Communication↗
PubMed1997

Finding association rules on heterogeneous genome data.

A novel approach for discovery of knowledge from genome data, which has been recently watched with interest in the research area of database, is applied to finding unified rules spreading over sequence, structure, and function of protein. As the result of experiments using data extracted from PDB, SWISS-PROT, and PROSITE, some association rules stating sequential/structural/functional aspects of two kinds of endopeptidases were found.

Computer Simulation↗
PubMed2026

Genetic modifiers of APOE-ε4-associated cognitive decline.

The APOE-ε4 allele is the strongest genetic risk factor for late-onset Alzheimer's disease. However, APOE-ε4 is not deterministic, highlighting the need to identify additional genetic and environmental factors. APOE-ε4 has been linked to accelerated cognitive decline, so we sought to investigate genetic factors that modify APOE-ε4-associated cognitive decline. We conduct cross-ancestry APOE-ε4-stratified and interaction GWAS using harmonized cognitive data from 32,778 participants, including 29,354 non-Hispanic White and 3,424 non-Hispanic Black individuals. Our primary outcome is late-life cognition, measured using harmonized composite scores for memory, executive function, and language, modeled as continuous traits reflecting both normative cognitive aging and disease-related decline. We identify two genome-wide significant loci in APOE-ε4 carriers, reaching genome-wide significance for executive function. These loci also demonstrate nominal associations across the other domains, suggesting broad effects on cognition. In non-carriers, we identify a genome-wide significant association at ITGB8 restricted to executive function, and another locus associated with language. We further link these loci to SEMA6D, GRIN3A, and ITGB8 through expression and methylation databases. Post-GWAS analyses implicate additional genes including SLCO1A2, and DNAH11. Genetic correlation analyses reveal differences by APOE-ε4 status for immune-related traits, suggesting immune-related predispositions may exacerbate cognitive risk in APOE-ε4 carriers.

Humans↗
PubMed2026

GBFN: A gated bimodal fusion network leveraging foundation model embeddings for cancer drug sensitivity prediction.

Despite recent progress in deep learning for cancer drug sensitivity prediction, many existing models still rely on task-specific representation learning or relatively simple multimodal fusion, which may limit their ability to capture complex drug-cell interactions. To address this issue, we developed GBFN, a gated bimodal fusion network for continuous IC50 prediction that integrates pretrained drug and cell-line representations. Specifically, drug embeddings were obtained from SMI-TED, whereas cell-line embeddings were derived from transcriptomic profiles using BulkFormer. These two modalities were then combined through a dimension-wise gated fusion module and used to predict IC50 values in matched drug-cell line pairs. On the CCLE-based benchmark, GBFN outperformed representative neural baselines, including GraphDRP, TGSA, and TransEDRP, and achieved the best overall performance, with an R² of 0.8714 and an RMSE of 0.8938. Moreover, ablation analysis showed that the model using drug features and cell-line expression data with gated fusion performed better than the corresponding model using direct concatenation, indicating that the improvement was associated with the fusion strategy rather than with the input modalities alone. In addition, cell-line expression data were more informative than mutation data in the present setting, and adding mutation data to the model using drug features and expression data did not further improve performance. Across major cancer types, GBFN maintained generally high cell-line-level predictive performance, and perturbation-based attribution identified biologically relevant transcriptomic programs in selected drug-cell line settings. Together, these findings support GBFN as a compact and effective framework for continuous drug response prediction.

Humans↗