National Biomedical Information Services

National Library Of Medicine

Investigators

Linked publications, trials & patents

Abstract

In FY2025, NLM’s biomedical information services expanded and enhanced access to data and information to accelerate biomedical research and health. Provided access to biomedical literature Indexed ~1M MEDLINE® citations with medical subject headings (MeSH®). Upgraded the machine learning algorithm that applies MeSH indexing to MEDLINE citations for improved accuracy and retrieval. Added 1.6M citations to PubMed®, now with 39.4+M citations to biomedical literature. Added ~1M articles to PubMed Central® (PMC) to provide free public access to 11+M full-text journal articles; 7+M of which are downloadable in machine-readable forms to support automated analysis through bulk download or at scale in a commercial cloud platform using machine learning and AI. Digitized ~536,000 pages from 4,400+ print volumes, and 121,000+ pages from archival collections, providing worldwide access through NLM’s Digital Collections. Enhanced metadata for items already in Digital Collections as part of an upgrade to a new repository platform, as well as for newly ingested materials toward the discoverability of print, electronic, audiovisual, and image resources in Digital Collections. Expanded and maintained other web materials by adding 1150+ resources focused on health topics including nutrition, diabetes, children’s health and well-being, disease prevention, and more, comprising 17M pages to NLM's web archive collections, currently with 15.7TB of data and 50,000+ web resources. Provided metadata to ~6,500 web resource titles, leading to increased discoverability of NLM’s 1.7M print, electronic, audiovisual, and image resources. Filled ~50,000 interlibrary loan (ILL) requests for NLM materials. Continued management of the DOCLINE system, which enables member libraries across the country to provide ILL services. Enabled ILL activities between ~1,700+ medical libraries to fulfill ~352,000 requests through the NLM DOCLINE ILL request and routing system. Provided trusted information for biomedical research and health care Registered 43,000+ new clinical research studies and added 7,500+ new results summaries to ClinicalTrials.gov, providing access to 550,000+ studies and 70,000+ results summaries. Launched a modernized ClinicalTrials.gov website and implemented a modernized Protocol Registration and Results System with continued improvements to user experienced based on their feedback. Making clinical trial information visible and available in ClinicalTrials.gov improves transparency, ensures accountability, and fosters public trust in science. Added 1.2B genetic sequences, including 35M bacterial sequence records, to the GenBank® database of all publicly available DNA sequences. Added 88M records to RefSeq, a database of reference sequences that includes genomic, transcript, and protein data, which generates comprehensive human genome annotation results. Added 1+M annotated human genetic sequence variants to ClinVar, which reports on relationships among human genome variations and human phenotypes. Improved ClinVar support for somatic cancer variants by using clinical impact classification standards. Provided 39+M Sequence Read Archive (SRA) data records in multiple formats, including an NLM-processed version (SRA normalized) and an NLM-processed version with a significantly reduced data storage footprint (SRA Lite), available on two commercial cloud platforms to enable the use of machine learning and AI. Making the data available in multiple formats facilitates more efficient access, retrieval, and use. Continued a multi-year effort to create a sustainable and extensible next generation data system capable of supporting the evolving needs of users into the future. This included developing a modernized cloud-based SRA application that enables NLM to more efficiently process a subset of submitted data in the most common format. Continued to lead the NIH Comparative Genomics Resource to facilitate comparative genomics analyses for eukaryotic organisms through an ecosystem of NLM repositories and community-created knowledgebases offering integrated cloud-ready data, tools, and interfaces compatible with community-provided organism resources. Released tools to enrich and improve the quality of genomic data submitted to GenBank to facilitate comparative analysis. Expanded the capabilities of a modern interface for finding and downloading genomic and taxonomic data at scale. Developed a web service to visualize and compare multiple annotated genomes at once. Processed genome sequence data for 500,000+ samples via the Pathogen Detection Pipeline to identify potential sources of human illnesses such as Salmonella, E. coli, and Listeria. Supported real-time foodborne pathogen surveillance by providing information that is used by the FDA, resulting in 1,330+ actions to protect consumers from foodborne illness and increase food safety. Provided access to antimicrobial resistance information for nearly 2.5M pathogens via AMRFinderPlus. Expanded the total number of human reference variants (with ‘RS’ unique identifiers) in the database of Single Nucleotide Polymorphisms (dbSNP) by adding 74+M new human genetic variants for a total of 1+B. Released a major update of the Allele Frequency Aggregator (ALFA) data collection, providing aggregated allele frequency data for ~1M variants with documented clinical significance – a 74% increase from the prior release – and established the ALFA data collection release 4 as the sole public allele frequence source for 27,000+ of these clinically relevant variants. In preparation for the next data collection release, dbSNP also processed 1+B new submitted variants (with ‘SS’ unique identifiers), including data from the All of Us Research Program, bringing the total to 5.6B. Improved the utility of the NIH Genetic Testing Registry® (GTR) by incorporating 1,800 new clinical genetic tests and updating information for 200 previously included tests, for a current total of 70,000 clinical genetic tests, 200 research genetic tests, and 50 microbe tests to more accurately reflect the current genetic testing market. Expanded DailyMed® drug labeling information for 154,000 drugs and added information for 340,000+ medical devices in the AccessGUDID database of than 4.9+M medical devices. Provided trusted health information to 585+M users who viewed 974+M MedlinePlus® pages and MedlinePlus Connect responded to 268+M electronic requests from health IT systems. Promoted the discovery of NIH investments in women’s health research with DiscoverWHR, a centralized resource for women’s health research and information from NIH that supports NIH-wide efforts to close the gaps in women’s health across the life course. DiscoverWHR had 58+K views from 26+K unique users, demonstrating the NIH's impact in the field of women's health research. Supported the development and use of health data standards Developed, maintained, and disseminated health data standards to facilitate the interoperability of health data to support federal health IT interoperability requirements and research. Expanded the SNOMED International Edition and the U.S. Extension with 7,880+ and 470+ concepts respectively. Added 610+ codes to LOINC and expanded the RxNorm® drug terminology by 320+ concepts. Enhanced the NIH Common Data Element (CDE) Repository to promote the adoption of CDEs in NIH-funded biomedical research. Updated repository functionality to allow users to browse CDEs by topic, to allow registered users to check their own submission files for accuracy and completeness, to encourage the reuse of existing CDEs, and to reduce the time required for quality assurance. Re-launched the enhanced Dataset Catalog for federated search of biomedical datasets from various repositories, driven by NLM’S Dataset Metadata Model (DATMM), increasing access to biomedical datasets 25-fold from 81K to almost 2M. The beta Dataset Catalog had 12,000+ users who conducted ~11,000 searches accessing ~4,000 separate datasets. Enhanced public awareness of NLM biomedical information services Increased public awareness of NLM resources by hosting 6 Michael E. DeBakey Fellows in the History of Medicine to support research using NLM historical collections; Ada Lovelace Computational Health Lectures on neuroscience, data science and AI for improving health, and the language of genomes; 3 History Talks promoting awareness and use of NLM’s historical collections for research; 41 online exhibitions; and 31 traveling exhibitions to share stories from NLM collections and raise global awareness of NLM’s products and services. NLM also co-sponsored the annual Joseph Leiter NLM/MLA Lectureship, which in 2025 featured the lecture Collaborative Intelligence and the Communication Imperative: Humans, Machines, and the Future of Impact. Supported generative AI pilot projects for improving workflow efficiencies and streamlining operations. For example, the NLM Dataset Catalog tested the use of generative AI to facilitate the normalization of disparate metadata from various data repositories for inclusion and discoverability. Results show the time required to ingest metadata decreased from ~100 hours to ~8 hours. In addition, evaluated the use of generative AI to improve accessibility of NLM resources by using a scanning engine to identify accessibility problems, then using an AI system to recommend improvements to resolve the identified problems.