Biomedical Text Mining and Ontologies

GLESKit: an MCP-based system for gene-level biomedical investigation and evidence synthesis

GLESKit is a gene-centered system that synthesizes biomedical evidence by integrating multiple data types, achieving high report quality and performance in benchmark tasks.

MutQA: A Cross-Validated Q&A Dataset for Genetic Mutations

The paper introduces MutQA, a cross-validated Q&A dataset linking genetic mutations to their functional consequences, and demonstrates its utility in improving LLM performance.

Phenologs at knowledge-graph scale: a mouse-to-slime-mold catalog of candidate genes for thousands of human diseases

The paper identifies 61,243 phenologs between human diseases and model organism phenotypes, using a hypergeometric method to rank candidate genes and validate their predictive power.

Extracting massive ecological data on state and interactions of species using large language models

This study uses large language models to extract ecological interactions from 83,910 scientific papers, creating a network of 144,402 interactions between 36,471 taxa with high precision and sensitivity.

Hybrid Semantic-Lexical Retrieval for Source-to-CDE Mapping in the NCI caDSR

The paper develops a hybrid retrieval system for mapping source data elements to CDEs, achieving high recall across multiple datasets.

Large language model classifies prior evidence in gene-phenotype associations

PESTO classifies gene-phenotype associations as Novel, Hypothesized, Existing, or Established using a large language model and structured genetic evidence, identifying 7% as Novel and 21% as Hypothesized in 940 associations.

The accessible human surfaceome

The paper curates an accessible human surfaceome using an LLM pipeline, outperforming existing databases and identifying new candidate surface proteins.

RDEmbed: A Sentence-Embedding Framework for Rare Disease Phenotype Representation and Phenotypic Similarity Analysis

RDEmbed uses sentence embeddings to analyze rare disease phenotypes from clinical text, outperforming ontology-based methods in similarity analysis and clustering.

Drug-centric knowledge graph: representing therapeutic modality, pharmacology and multiscale biology

The paper introduces a drug-centric knowledge graph that organizes therapeutic interventions, linking them to biological entities and pharmacological properties through a structured schema.

PatientPheX: A Full-Text Biomedical Corpus for Patient-Specific Phenotype Information Extraction

PatientPheX is a biomedical corpus for extracting and normalizing patient-specific phenotype information from full-text articles, achieving high entity and relation extraction performance.

An Expert-Curated Benchmark for Frontier and Local LLMs Querying a Biomedical Knowledge Graph

The paper benchmarks frontier and local LLMs on a biomedical knowledge graph using expert-curated questions and a new evaluation metric called psi.

KRAKEN: A provenance-tracked knowledge graph for multiomic and wellness research

A Biolink-compliant graph (∼15 million nodes, ∼113 million edges) adds RefMet and LIPID MAPS metabolite and lipid classifications, plus polygenic score and clinical data element entities, to Translator content.

AXIS: Agentic Extraction and Integration of Simulation Scenarios for Quantitative Systems Pharmacology

AXIS converts published QSP models and their companion publications into executable simulation configurations by integrating entity grounding, scenario extraction, and parameter assignment.

Validated Semantic Scoring Improves Filtering of Generic Concepts from Knowledge-Graphs for Drug Repurposing Queries

This paper introduces a semantic precision score to filter generic concepts from biomedical knowledge graphs, outperforming existing methods in drug repurposing queries.

ddkg.skill: A Compositional Agent Skill for Translating Biomedical and Bioinformatics Questions into Cypher for the Data Distillery Knowledge Graph

ddkg.skill translates biomedical questions into Cypher queries for the DDKG, improving accuracy through structured references and validation.

AI-Based Comparison of UniProt annotations and Literature Across Human Kinases

The paper uses AI to compare biological information from UniProt records and their cited literature for human kinases, finding literature provides better coverage for certain topics.

A low-annotation-budget PubMedBERT classifier for chondrogenesis regulator discovery via active learning

The paper develops a low-annotation-budget classifier using active learning and PubMedBERT to identify chondrogenesis regulators from PubMed abstracts, achieving high performance with fewer annotations.

A Comparative Benchmark of Biomedical Language Models for Concept Normalization from Real-World Text

This study benchmarks biomedical language models for concept normalization from real-world text, finding Qwen 2 Instruct improves accuracy when combined with SapBERT.

BELL: Biomodel Evidence and LLM-based Logic

BELL automates evidence retrieval and scoring for biological interactions, achieving high confidence in 49.5% of tested interactions and guiding curators effectively.

Corpus-wide causality: Algorithm design & application for aggregating gene-disease causal evidence

This paper introduces a method to compute a Corpus-Wide Causal Score for gene-disease pairs by integrating network-based and literature-based evidence, outperforming LLMs in causal classification.

A heterogeneous biomedical knowledge network framework for rare disease drug candidate prioritization: integrating Orphadata and DisGeNET via gene-bridge harmonization

This paper develops a network framework to prioritize drug candidates for rare diseases by integrating Orphadata and DisGeNET through gene-bridge harmonization and using a Graph Attention Network for node representation learning.

AI-Supported Extraction of Functional Tissue Unit Properties for Human Reference Atlas Construction

The paper introduces an AI framework to automate the extraction of functional tissue unit properties from publications, supporting the construction of the Human Reference Atlas.

AI semantics for biomedical data integration

This paper describes an AI semantics workflow using LLM agents to integrate multimodal biomedical data through ontologies and knowledge graphs.

An annotation-overlap-flagged rare-disease gene-prioritisation benchmark and PMC index recipe

The paper introduces a benchmark for rare-disease gene prioritization using 1,047 cases and a recipe for building a PMC article index, without comparing tools.

Automatic bioinformatic software named entity recognition from literature

SNAIL is a named entity recognition framework that identifies bioinformatics software and databases in scientific texts, outperforming existing methods and enabling meta-analysis of tool usage.

Every Cure Knowledge Graph: A Unified Biomedical Knowledge Graph for Drug Repurposing

The paper introduces EC-KG, a unified biomedical knowledge graph for drug repurposing that integrates multiple existing knowledge graphs to enhance coverage and uncover novel mechanistic pathways.

Searching the Druggable Genome using Large Language Models

The paper develops a server enabling large language models to access drug-gene interaction data through natural language queries.

Assessing the Reliability of LLM-Generated Phenotype-Genotype Associations Through External Validation

The paper evaluates LLM-generated phenotype-genotype associations using external validation, finding 74.19% match genomic knowledge bases with varying support levels.

Automating Biomedical Knowledge Graph Construction For Context-Aware Scientific Inference

AutoBioKG constructs context-aware biomedical knowledge graphs to improve scientific inference by encoding environmental conditions and entity attributes, outperforming baselines in multiple benchmarks.

GlycoMeSH: linking glycan structures to biomedical context for systematic enrichment analysis

GlycoMeSH links glycan structures to biomedical contexts through a database and enrichment workflow, improving association recovery and coverage over existing methods.