key: cord-0002078-a564l6vs
authors: nan
title: Database resources of the National Center for Biotechnology Information
date: 2016-01-04
journal: Nucleic Acids Res
DOI: 10.1093/nar/gkv1290
sha: 819902e2222b531ae0a6016763e43618b2988516
doc_id: 2078
cord_uid: a564l6vs

The National Center for Biotechnology Information (NCBI) provides a large suite of online resources for biological information and data, including the GenBank(®) nucleic acid sequence database and the PubMed database of citations and abstracts for published life science journals. Additional NCBI resources focus on literature (PubMed Central (PMC), Bookshelf and PubReader), health (ClinVar, dbGaP, dbMHC, the Genetic Testing Registry, HIV-1/Human Protein Interaction Database and MedGen), genomes (BioProject, Assembly, Genome, BioSample, dbSNP, dbVar, Epigenomics, the Map Viewer, Nucleotide, Probe, RefSeq, Sequence Read Archive, the Taxonomy Browser and the Trace Archive), genes (Gene, Gene Expression Omnibus (GEO), HomoloGene, PopSet and UniGene), proteins (Protein, the Conserved Domain Database (CDD), COBALT, Conserved Domain Architecture Retrieval Tool (CDART), the Molecular Modeling Database (MMDB) and Protein Clusters) and chemicals (Biosystems and the PubChem suite of small molecule databases). The Entrez system provides search and retrieval operations for most of these databases. Augmenting many of the web applications are custom implementations of the BLAST program optimized to search specialized datasets. All of these resources can be accessed through the NCBI home page at www.ncbi.nlm.nih.gov.

The National Center for Biotechnology Information (NCBI) at the National Institutes of Health was created in 1988 to develop information systems for molecular biology. In addition to maintaining the GenBank R (1) nucleic acid sequence database, which receives data through an inter-national collaboration with the DNA Data Bank of Japan (DDBJ) and the European Nucleotide Archive (ENA) as well as from the scientific community, NCBI provides many other kinds of biological data as well as retrieval systems and computational resources for the analysis of GenBank and other data. This article provides a summary of recent developments, including both new and updated resources, followed by an introduction to the Entrez system and a brief review of the suite of NCBI resources. All resources discussed are available through the NCBI home page at www.ncbi.nlm.nih.gov and can also be located using the NCBI Web Site database available in Entrez search menus. In most cases, the data underlying these resources and executables for the software described are available for download at ftp.ncbi.nlm.nih.gov.

PubMed Labs is an NCBI initiative for developing experimental projects by involving the user community throughout the process. Projects in PubMed Labs may be early versions of new services, proposed features for existing resources, or novel database content. PubMed Labs projects will be described on the NCBI Insights blog (ncbiinsights.ncbi.nlm.nih.gov) in a new category of posts labeled PubMed Labs. Each post will describe the project and provide instructions for how users can test its functions, and will also indicate what results to expect. We encourage users to share their experiences with us by commenting on these posts. Currently there are two initial projects in PubMed Labs: SmartBLAST and PubMed Also-Viewed.

SmartBLAST is an experimental tool that makes it easier to accomplish common protein sequence analysis tasks such as finding a candidate name for a protein, identifying highly conserved regions and locating segments that are present in closely related database sequences but that are missing from the query. SmartBLAST does this by performing two parallel BLASTp searches: one that retrieves the closest matching sequences available, and another that finds To the closest matches from well-annotated sequences from model organisms. The tool then constructs a multiple sequence alignment between the query and five of the closest matches, and displays this as a phylogenetic tree. Smart-BLAST accomplishes all of this in much less time than it takes to run a typical BLASTp search. Links to Smart-BLAST may be found on the main BLAST page as well as on BLASTp results pages.

PubMed Also-Viewed is a link on some PubMed abstract pages that shows PubMed records that other users have viewed with the current record. When present, the link is labeled Articles frequently viewed together and appears on the right side of the abstract page. At the time of this writing the link was available for about 5% of PubMed records.

In 2014 NCBI revised the main home page to include six prominent buttons that lead to pages focused on a particular set of services. The Submit button loads a page that helps submitters choose the correct mechanism(s) for submitting data to NCBI, while the Download page provides access to the FTP site and related tools. The Learn page introduces users to a variety of NCBI educational resources and programs, while the Analyze page leads to software tools developed by NCBI. Intended for software developers, the Develop page links to the several NCBI APIs and software toolkits, as well as to the NCBI GitHub repository. Finally, the Research page allows users to explore NCBI research projects and collaborations, along with NCBI research groups and associated staff. Most of these pages also feature a Feedback button so that users can easily provide comments and suggestions.

A new addition to the Database of Genotypes and Phenotypes (dbGaP) is the dbGaP Data Browser (DDB; https: //www.ncbi.nlm.nih.gov/gap/ddb/) that enables users to explore variant calls, genotype calls and supporting sequence read alignments for controlled access datasets in a genomic context. It is a companion to the dbGaP Beacon resource and provides a graphical interface for reviewing its results. The interface is modeled after the 1000 Genomes project browser; however, its content represents a combination of data from the dbGaP controlled access and public databases. Depending upon authentication credentials, users can access read-only views of selected datasets or download datasets to which they have been granted access. Within the browser, the Subjects widget supports faceted and free-text searching for browser tracks of Sequence Read Archive (SRA) alignments for samples of interest. Sample attributes such as the study accession, BioSample ID, sex, population, tumor/normal status and the availability of sample genotype data are just a few of the available search facets. The browser's graphical display allows users to view the read alignments from selected study participants alongside data from ClinVar, the Database of Single Nucleotide Polymorphisms (dbSNP) and Gene. The Genotypes table in the browser provides access to individual-level and aggregate General Research Use (GRU) dataset genotypes for studies having samples whose SRA alignments are displayed in the graphical view. Within the table, users can view frequencies or counts for either alleles or samples. DDB also includes functions found in other NCBI browsers, including feature search and support for user data uploads. Further documentation for DDB is available online (https: //www.ncbi.nlm.nih.gov/gap/ddb/help/).

The NCBI Pathogen Detection project (www.ncbi.nlm.nih. gov/pathogens/) is a new system that facilitates real-time surveillance of bacterial pathogens and foodborne disease. This project is a collaboration with several public health agencies including the Centers for Disease Control (CDC), the Food and Drug Administration (FDA), the United States Department of Agriculture Food Safety and Inspection Service (USDA-FSIS), Public Health England (PHE) and several state and regional labs in the US. Samples collected by these agencies are sequenced and submitted to NCBI where an automated pipeline clusters and identifies the sequences and then quickly reports the results back. Collecting and analyzing pathogen sequence data obtained from food, the environment and human patients reveals potential sources of contamination and facilitates traceback investigations and responses to outbreaks. Currently the project focuses on the following bacterial groups: Campylobacter, Listeria, Salmonella and the combination of Escherichia coli and Shigella. Analysis results are available from the NCBI FTP site (ftp.ncbi.nlm.nih.gov/pathogen/).

In response to user feedback, minor changes were made to the PubMed interface so that popular features are easier to find. The Related citations feature was renamed to Similar articles (the algorithm to generate the results was not modified). The Save search and RSS links that allow users to create My NCBI automatic email alerts were renamed to Create alert and Create RSS. Additionally, to provide a less-cluttered PubMed results display, the status tag lines (e.g. [PubMed--as supplied by publisher]) were removed from the summary results. The abstract display was also enhanced to show the transliterated title for citations originally published in a non-English language that do not include an English title.

PubMed Mobile was updated with a number of styling modifications and enhancements including a Trending articles feature on the home page. PubMed Mobile summary results now include the Related searches discovery tool as well as sorting options and additional filter selections. These Discovery tools appear below the results on mobile devices with smaller screen sizes. A Show full citation link was added to streamline the PubMed Mobile abstract page, and clicking this link displays the citation, author(s) and affiliation details. Clicking a linked author name in the resulting list displays a sorted set of citations for that author.

In response to requests from authors and users, PubMed Central (PMC) added several new features in the last year.

Nucleic Acids Research, 2016, Vol. 44, Database issue D9 Among these features is the new citation exporter that makes it easy to retrieve pre-formatted citations in AMA, MLA or APA styles that users can then copy, paste or download into bibliographic reference manager software. Additionally, in order to facilitate text and data mining for articles in the Open Access Subset, PMC is now providing plain text files for these articles on the PMC FTP site. The files contain the full text of the article, extracted either from the XML or PDF source files.

PMC is also now serving as the public access repository for a number of federal agencies in addition to NIH that support scientific research. As of January 2015, the Centers for Disease Control and Prevention (CDC), the National Institute of Standards and Technology (NIST) and the Department of Veterans Affairs (VA) have implemented public access policies requiring researchers who are supported by these agencies to make the resulting manuscripts publicly available in PMC within 12 months of publication. The NIH manuscript submission (NIHMS) system has been extended to support researchers from these additional agencies.

My NCBI (www.ncbi.nlm.nih.gov/account/) provides a specialized biosketch tool called SciENcv (Science Experts Network Curriculum Vitae) for users who wish to create, download and share biosketches for NIH grant applications. SciENcv was updated to support the new NIH Biographical Sketch format that became mandatory for grant applications with due dates of May 2015 or later. SciENcv can upgrade biosketches stored in the previous NIH format to the new version and it also supports the NSF biosketch format on an alpha-release basis. SciENcv is integrated with the My Bibliography citation tool, which is required for NIH extramural grantees to demonstrate compliance with the NIH Public Access Policy. Among other improvements to the user interface, the process used to describe scientific accomplishments has been enhanced to allow easier import of citations.

GTR. The Genetic Testing Registry (GTR) collects and displays information that has been submitted by providers about their genetic tests (2) . GTR accepts submissions for germline, somatic and research tests. Submission formats have been expanded from interactive wizards and Excel templates to include submission of tests as XML files. Submitted data appear on the GTR web site within 24-48 h (http://www.ncbi.nlm.nih.gov/gtr/docs/fulltest/). GTR submitters are required to review their laboratory and test data annually (www.ncbi.nlm.nih.gov/gtr/docs/annual review/). To support automated test updates, submitters can download all of their submitted test data to an Excel template file to edit and upload. Records that have not been reviewed within a year of the previous review are marked out of date on the GTR site, and records that have not been reviewed in two years will be removed from display. This stringency in representing current information may result in differences in test counts between GTR and other public websites.

In addition to providing an advanced search to find tests by a variety of attributes, GTR now provides an All GTR display that not only allows users to interrogate content about tests, conditions, genes and laboratories simultaneously, but also organizes the data in each domain to show high-value information. For example, the Tests tab shows the name of the test, the name and location of the lab and the methods used, along with links to view details about conditions and test targets. The Conditions tab lists symptoms to support recognition and links to all tests, genes and GeneReviews for each condition. The Genes tab displays associated conditions for each gene and provides links to tests for each gene. Each of these tabs also allows users to select multiple items (e.g. genes, conditions, laboratories) and retrieve associated tests.

NCBI has continued to revise the genome area of the NCBI FTP site (ftp.ncbi.nlm.nih.gov/genomes/). The new directories within the genome area (genbank, refseq and all) now provide a standard set of data files for over 54 000 assemblies. The genbank directory contains data submitted directly to GenBank (or an INSDC database), while the refseq directory contains data that are part of the Reference Sequence (RefSeq) project. The common data unit within these three directories is a subdirectory corresponding to a record in the Assembly database (3) and given a name consisting of the Assembly accession followed by the name of the assembly. For example, the human GRCh38 release has assembly accession GCF 000001405.26, and so the subdirectory is named GCF 000001405.26 GRCh38. Users are encouraged to search the Assembly database directly to find these accessions, assembly names and other details about the datasets. The genbank and refseq directories collect the Assembly subdirectories within broad taxonomic directories (e.g. plant, bacteria and vertebrate mammalian) and also directories for each species. Each Assembly subdirectory contains a standard set of files including FASTA and GenBank/GenPept data for genome, transcript and protein sequences, along with GFF3 and feature table files for annotated genome records. Older directories in the genome FTP area that are no longer being updated will be moved to an archive area by late 2015.

In response to the rapidly growing number of prokaryotic genomes being submitted to NCBI, the scope of the Gene database changed in 2014. Going forward, NCBI will create Gene records only for reference and selected representative genomes for a single prokaryotic species (www. ncbi.nlm.nih.gov/refseq/about/prokaryotes/). Gene records for strains not included in these sets are being discontinued, and their record pages will contain messages providing more detail about each case (www.ncbi.nlm.nih.gov/refseq/ about/prokaryotes/faq/).

Protein database records now have a prominent link to the Identical Protein Report at the top of the record page. This report displays the accessions of all other protein records whose sequences are identical to that of a given protein.

The report also provides links to the CDS sequence in Nucleotide for each protein. While this report is available for all proteins, it is especially useful for the non-redundant RefSeq WP sequences introduced in 2013 (4). Because many WP sequences represent a set of identical proteins that may not have separate species-or strain-specific records, these WP sequences are connected to not one but a corresponding set of Nucleotide CDS sequences. The Identical Protein Report clarifies these relationships. For example, WP 002317106 collects over 40 thioredoxin sequences from several species and strains of Enterococcus. The report is also available through the E-utility EFetch with &rettype = ipg.

A new search box on the BLAST home page makes it easy to find a genomic BLAST search page for a given organism. The box has an autocomplete feature that presents suggestions when a user starts to type. Choosing an organism loads a BLAST search page with the best genomic database for that organism preselected. The search box also produces metagenomic and microbial suggestions.

MOLE-BLAST is a new tool that classifies multiple nucleotide query sequences and displays their relationships. Ideal input for this tool is a set of sequences representing a specific locus from a group of organisms rather than the entire genome of an organism or a set of unannotated contigs. Example input would be 16S sequences from different bacteria or ITS sequences from fungi. MOLE-BLAST first assigns each query in the input set to a cluster using BLAST, thereby grouping the queries by locus. Second, it performs a database search to find top matches for each query. Third, it computes a multiple alignment (using MUSCLE) between the queries and their top matches, and presents this analysis as a phylogenetic tree.

Both the PubChem Compound and Substance record view pages were completely redesigned in the past year. The new pages use a responsive design approach that optimizes the display on a variety of screen sizes, including both touchand mouse-based interfaces. These reports include an improved and expanded table of contents that makes navigation easier and users can now bookmark particular sections of the reports. Full details of the many improvements are discussed in posts on the PubChem blog (pubchemblog.ncbi.nlm.nih.gov).

Entrez (5) is an integrated database retrieval system that provides access to a diverse set of 39 databases that together contain 1.7 billion records (Table 1 ). Links to the web portal for each of these databases are provided on the Entrez GQuery page (www.ncbi.nlm.nih.gov/gquery/). Entrez supports text searching using simple Boolean queries, downloading of data in various formats and linking of records between databases based on asserted relationships. In their simplest form, these links may be cross-references between a sequence and the abstract of the paper in which it is reported, or between a protein sequence and either its coding DNA sequence or its 3D-structure. Computationally derived links between neighboring records, such as those based on computed similarities among sequences or among PubMed abstracts, allow rapid access to groups of related records. Several popular links are displayed as Discovery Components in the right column of Entrez search result or record view pages, making these connections easier to find and explore. The LinkOut service expands the range of links to include external resources, such as organismspecific genome databases. The records retrieved by Entrez can be displayed in many formats and downloaded singly or in batches.

NCBI receives data from three sources: direct submissions from external investigators, national and international collaborations or agreements with data providers and research consortia, and internal curation efforts. The Data Source column in Table 1 indicates those mechanisms by which each Entrez database receives data. More information about the various collaborations, agreements and curation efforts are available through the home pages of the individual resources.

The Entrez Programming Utilities (E-Utilities) constitute the Application Programming Interface (API) for the Entrez system. The API includes nine programs that support a uniform set of parameters used to search, link and download data from the Entrez databases. EInfo provides basic statistics on a given database, including the last update date, along with lists of all search fields and available links. ESearch returns the identifiers of records that match an Entrez text query and when combined with EFetch or ESummary, provides a mechanism for downloading the corresponding data records. ELink gives users access to the vast array of links within Entrez so that data related to an input set can be retrieved. By assembling URL calls to the E-utilities within simple scripts, users can create powerful applications to automate Entrez functions to accomplish batch tasks that are impractical using web browsers. Detailed documentation for using the E-Utilities is available at eutils.ncbi.nlm.nih.gov.

Entrez Direct is a set of executables that provides an interface to the E-utilities on the UNIX command line. These executables are designed so that the output of one can be passed directly as input to another using the UNIX pipe ('|'). In this way, it is straightforward to implement a diverse assortment of workflows. Entrez Direct also offers a utility named xtract that parses the XML output of ESummary and EFetch calls so that individual fields within records can be retrieved and formatted into custom tables, especially when combined with standard UNIX commands such as grep, sort, cut, awk or sed. Complex workflows can be conveniently saved as shell scripts for sharing or use 

The PubMed database contains citations from life science journals, many of which include abstracts and links to their full text articles.

PubMed Commons enables the community to share information and opinions on scientific publications. 

The MeSH database (7) includes information about the NLM controlled vocabulary thesaurus used for indexing PubMed citations, and provides an interface for constructing PubMed queries using MeSH terms.

The NCBI Bookshelf is an online service of the National Library of Medicine Literature Archive (NLM LitArch) that provides free access to the full text of books, reports, databases and documentation in the life sciences and health care fields.

ClinVar supports users who want to determine what has been reported about the medical relevance of human sequence variation (8) . 

MedGen organizes information about human disorders that have a genetic component (www.ncbi.nlm.nih.gov/ books/NBK159970/). Starting from freely available content in the semi-annual releases from UMLS (www.nlm.nih.gov/ research/umls/), MedGen adds recent content from OMIM, terms and relationships from the Human Phenotype (www.human-phenotype-ontology.org/) and ORDO (www. orphadata.org/cgi-bin/inc/ordo orphanet.inc.php) ontologies, and terms submitted from ClinVar and GTR. MedGen organizes terms from multiple sources by assigning them a concept ID, and then adds value by reporting practice guidelines, related genes from the Gene database, variants in ClinVar and available tests in GTR. MedGen supports querying for disorders that share clinical features as well as drugs and their responses. MedGen data can be downloaded using FTP as pipe-delimited (RFF) or CSV text files and using the E-utilities.

The Database of Genotypes and Phenotypes (dbGaP) (10) archives, distributes and supports submission of data that correlate genomic characteristics with observable traits. This database is a designated NIH repository for NIHfunded genome-wide association study (GWAS) results (grants.nih.gov/grants/gwas/). To protect the confidentiality of study subjects, dbGaP accepts only de-identified data and requires investigators to go through an authorization process in order to access individual-level data. Study documents, protocols and subject questionnaires are available without restriction.

PubMed Health provides information for consumers and clinicians about the prevention and treatment of diseases and conditions. The database specializes in reviews of clinical effectiveness research, containing both summaries for consumers and full technical reports.

NCBI maintains three databases for routine clinical applications: dbMHC, dbLRC and dbRBC. dbMHC focuses on the Major Histocompatibility Complex (MHC) and contains sequences and frequency distributions for MHC alleles, as well as genotype and clinical outcome information on hematopoietic cell transplants performed worldwide. dbLRC offers a comprehensive collection of alleles of the Leukocyte Receptor Complex (LRC) with an emphasis on KIR genes. dbRBC provides data on genes for Red Blood Cell (RBC) antigens along with access to the International Society of Blood Transfusion allele nomenclature of blood group alleles. dbRBC also hosts the Blood Group Antigen Gene Mutation Database (11) and integrates it with resources at NCBI. All three databases provide multiple sequence alignments, analysis tools to interpret homozygous or heterozygous sequencing results (12) and tools for DNA probe alignments.

The BioProject database is a central access point for metadata about research projects whose data are deposited in an INSDC database. BioProject provides links to the primary data from these projects, which range from focused genome sequencing projects to large international collaborations. These larger projects may have multiple sub-projects incorporating experiments that produce nucleotide sequence sets, Nucleic Acids Research, 2016, Vol. 44, Database issue D13 genotype/phenotype data, sequence variants or epigenetic information.

The Assembly database (3) collects metadata about genome assemblies that were either submitted to GenBank (or an INSDC database) or that are part of the RefSeq database. Assembly records also provide statistics about the genome as well as links to the sequence data in Entrez or in the genomes area of the FTP site.

The Genome database collects genomic sequencing projects for a given species and provides links to corresponding records in BioProject, Assembly, Nucleotide and Protein. Genome records collect genome assembly data at various levels of completion, ranging from genomes represented by scaffolds or contigs to fully assembled chromosomes with annotation. NCBI creates a Genome record for an organism if at least one assembly is available for that organism in the Assembly database. The Genome home page also provides links to an organism browser that lists the current status of all genomes annotated at NCBI.

The RefSeq database (13) is a non-redundant set of curated and computationally derived sequences for transcripts, proteins and genomic regions. RefSeq DNA and RNA sequences can be searched and retrieved from the Nucleotide database and the complete RefSeq collection is available in the RefSeq directory on the NCBI FTP site.

GenBank (1) is the primary nucleotide sequence archive at NCBI and is a member of the International Nucleotide Sequence Database Collaboration (INSDC). Sequences from GenBank are available from three Entrez databases: Nucleotide, EST and GSS (specified as nuccore, nucest and nucgss within the E-utilities). The Nucleotide database contains all GenBank sequences except those within the EST or GSS GenBank divisions. The database also contains WGS sequences, Third Party Annotation (TPA) sequences and sequences imported from the Structure database.

The PopSet database is a collection of related sequences and alignments derived from population, phylogenetic, mutation and ecosystem studies that have been submitted to GenBank. When available, PopSet alignments are shown in an embedded viewer on the PopSet record page.

SRA (14) is a repository for raw sequence reads and alignments generated by high-throughput nucleic acid sequencers. Data are deposited into SRA as supporting evidence for a wide range of study types, including de novo genome assemblies, genome wide association studies, single nucleotide polymorphism and structural variation analysis, pathogen identification, transcript assembly, metagenomic community profiling and epigenetics studies.

The Trace Archive contains sequence traces from gel and capillary electrophoresis sequencers. These data arise from whole genomes of pathogens, organismal shotgun and BAC clone projects, and EST libraries. A companion resource, the Trace Assembly Archive, contains placements of individual trace reads on a GenBank sequence.

CloneDB is a resource for finding descriptions, sources, map positions and distributor information about available clones and libraries (15) . For both genomic and cell-based clones and libraries, CloneDB contains information about the sequences themselves, such as their genomic mapping positions and associated markers, along with details about how the libraries were constructed.

The Probe database is a registry of nucleic acid reagents designed for use in a wide variety of biomedical research applications including genotyping, SNP discovery, gene expression, gene silencing and gene mapping. Probe also includes information on reagent distributors, probe effectiveness and computed sequence similarities.

The BioSample database provides annotation for biological samples used in a variety of studies submitted to NCBI, including genomic sequencing, microarrays, genome wide association studies (GWAS) and epigenomics (16) . The database promotes the use of structured and consistent attribute names and values that describe what the samples are as well as information about their provenance, where appropriate.

The NCBI taxonomy database is a central organizing principle for the Entrez biological databases and provides links to all data for each taxonomic node, from superkingdoms to subspecies (17) . The taxonomy database reflects sequence data from virtually all of the formally described species of prokaryotes and about 10% of the eukaryotes. The Taxonomy Browser can be used to view the taxonomy tree or retrieve data from any of the Entrez databases for a particular organism or group. In 2013 the Taxonomy database began including type material for prokaryotic type strains and eukaryotic type specimens (18) . As of January 2014 NCBI no longer assigns taxonomy IDs to bacterial strains that do not already have taxonomy IDs (19) . Instead, such sequences will be assigned the taxonomy ID of the bacterial species, while the strain will be included in the source information D14 Nucleic Acids Research, 2016, Vol. 44, Database issue of the sequence record. In addition, the sequence record will be linked to a BioSample record that will contain strain information such as relevant culture collections and details about how the strain was isolated.

The Genome Reference Consortium (GRC) (www.genomereference.org) is an international collaboration between the Wellcome Trust Sanger Institute, the Genome Institute at Washington University, EMBL and NCBI. The GRC aims to produce assemblies of higher eukaryotic genomes that best reflect complex allelic diversity that is consistent with currently available data. The GRC currently produces assemblies for human (GRCh38), mouse (GRCm38) and zebrafish (GRCz10). Between major assembly releases the GRC provides minor patch releases that provide additional sequence scaffolds that either correct errors in the assembly (fix patches) or add an alternate loci (novel patches). GRC staff then incorporate these changes into the next major assembly release. GRC data are available for download from the NCBI FTP site (ftp.ncbi.nlm.nih.gov/pub/grc/) and the new genomes FTP area (see above).

The Database of Single Nucleotide Polymorphisms (db-SNP) (20) is a repository of all types of short genetic variations <50 bp in length, and so is a complement to dbVar (see below). dbSNP accepts submissions of both common and polymorphic variations, and contains both germline and somatic variations. In addition to archiving molecular details for each submission and calculating submitted variant locations on each genome assembly, dbSNP maintains information about population-specific allele frequencies and genotypes, reports the validation state of each variant and indicates if a variation call may be suspect because of paralogy (21) .

The Database of Genomic Structural Variation (dbVar) is an archive of large-scale genomic variants (generally >50 bp) such as insertions, deletions, translocations and inversions (22) . These data are derived from several methods including computational sequence analysis and microarray experiments.

The Variation Viewer (www.ncbi.nlm.nih.gov/variation/ view) displays human variations from dbSNP, dbVar and ClinVar in the context of the current and previous human reference genome assemblies, now GRCh38 and GRCh37 (23) . Variation Viewer provides users with genome-wide access to variations, both graphically and in tabular format. Users can search for variants by gene symbol, variant ID, or chromosomal coordinates, and can also upload their own data in several popular formats.

The Virus Variation Resource is an outgrowth of the Influenza virus and Dengue virus resources and has been updated to include West Nile virus, Middle Eastern Respiratory (MERS) coronavirus and Ebolavirus (23) (24) (25) (26) . The resource employs computational pipelines and manual curation to create consistent sequence annotations and metadata vocabularies across all sequences from constituent viruses. These standardized data are leveraged by a specialized search interface and a suite of tools designed to support the retrieval and display of large virus sequence datasets.

The Epigenomics database collects data from studies examining epigenetic features such as post-translational modifications of histone proteins, genomic DNA methylation, chromatin organization and the expression of non-coding regulatory RNA (27) . The Epigenomics database provides displays (genome tracks) of the raw data (stored in the Gene expression omnibus (GEO) and SRA databases) mapped to genomic coordinates. Data from the Roadmap Epigenomics project, currently stored in GEO (www.ncbi.nlm.nih.gov/ geo/roadmap/epigenomics/), are being mirrored and are available for viewing and downloading.

Gene (28) provides an interface to curated sequences and descriptive information about genes with links to a wide variety of gene-related resources. These data are accumulated and maintained through several international collaborations in addition to curation by NCBI staff. The complete Gene dataset, as well as organism-specific subsets, is available in the compact NCBI Abstract Syntax Notation One (ASN.1) format on the NCBI FTP site. The gene2xml tool converts the native Gene ASN.1 format into XML and is available at ftp.ncbi.nlm.nih.gov/toolbox/ncbi tools/ converters/by program/gene2xml/.

As part of the Locus Reference Genomic (LRG) collaboration (www.lrg-sequence.org), RefSeqGene provides stable, standard human genomic sequences annotated with mR-NAs for well-characterized human genes (13) . RefSeqGene records are part of the RefSeq collection and are used to establish numbering systems for exons and introns and for reporting and identifying genomic variants, especially those of clinical importance (29) . RefSeqGene records can be retrieved from the Nucleotide database using the query 'refseqgene[keyword]', are available on corresponding Gene reports and can be downloaded from ftp.ncbi.nlm.nih.gov/ refseq/H sapiens/RefSeqGene.

The conserved CDS database (CCDS) project is a collaborative effort between NCBI, the European Bioinformatics Institute, the Wellcome Trust Sanger Institute (WTSI) Nucleic Acids Research, 2016, Vol. 44, Database issue D15 and the University of California, Santa Cruz (UCSC). The CCDS compiles a set of human and mouse protein coding regions that are consistently annotated and of high quality (30) . The collaborators prepare the CCDS set by comparing the annotations they have independently determined and then identifying those coding regions that have identical coordinates on the genome. Those regions that pass quality evaluations are then added to the CCDS set. The CCDS sequence data are available at ftp.ncbi.nlm.nih.gov/ pub/CCDS/.

GEO (31) is a data repository and retrieval system for highthroughput functional genomic data generated by microarray and next-generation sequencing technologies. In addition to gene expression data, GEO accepts data from studies of genome copy number variation, genome-protein interaction surveys and methylation profiling studies. The repository can capture fully annotated raw and processed data, enabling compliance with reporting standards such as 'Minimum Information About a Microarray Experiment' (MIAME) (23, 24) . GEO data are housed in two Entrez databases: GEO Profiles, which contains quantitative gene expression measurements for one gene across an experiment, and GEO DataSets, which contains entire experiments.

UniGene (32) is a system for partitioning transcript sequences (including ESTs) from GenBank into a nonredundant set of clusters, each of which contains sequences that seem to be produced by the same transcription locus. UniGene clusters are created for all organisms for which there are 70 000 or more ESTs in GenBank.

HomoloGene is a system that automatically detects homologs, including paralogs and orthologs, among the genes of 21 completely sequenced eukaryotic genomes. Homolo-Gene reports include homology and phenotype information drawn from Online Mendelian Inheritance in Man (OMIM) (33) , Mouse Genome Informatics (MGI) (34), the Zebrafish Information Network (ZFIN) (35) , the Saccharomyces Genome Database (SGD) (36) and FlyBase (37) . Information about the HomoloGene build procedure is provided at www.ncbi.nlm.nih.gov/HomoloGene/HTML/ homologene buildproc.html.

In addition to genomic and transcript sequences, the RefSeq database (13) contains protein sequences that are curated and computationally derived from these DNA and RNA sequences. RefSeq protein sequences can be searched and retrieved from the Protein database, and the complete RefSeq collection is available in the RefSeq directory on the NCBI FTP site.

As part of standard submission procedures, NCBI produces conceptual translations for any sequence in Gen-Bank that contains a coding sequence and places these protein sequences in the Protein database. In addition to these GenPept sequences, the Protein database also contains sequences from TPA, UniProtKB/Swiss-Prot (38) , the Protein Research Foundation (PRF) and the Protein Data Bank (PDB) (39) .

Molecular modeling database (MMDB) (40) contains experimentally determined coordinate sets from PDB (39) augmented with domain annotations and links to relevant literature, protein and nucleotide sequences, chemicals (PDB heterogens), and conserved domains in the Conserved Domain Database (CDD) (41) . MMDB also provides interactive views of the data in Cn3D (42), the NCBI structure and alignment viewer. MMDB provides structural neighbors for each record based on similarities computed by the VAST algorithm between compact structural domains within protein structures (43, 44) .

Score Matrices representing domains taken from the Simple Modular Architecture Research Tool (Smart) (46) , Pfam (47) , TIGRFAM (48) and from domain alignments derived from the Clusters of Orthologous Groups (COGs) database and the Protein Clusters database. In addition, CDD includes superfamily records that contain sets of CDs from one or more source databases that generate overlapping annotation on the same protein sequences.

The Protein Clusters database contains sets of almost identical RefSeq proteins encoded by complete genomes from prokaryotes, eukaryotic organelles (mitochondria and chloroplasts), viruses and plasmids, as well as from some protozoans and plants. The clusters are organized in a taxonomic hierarchy and are created based on reciprocal besthit protein BLAST scores (49) .

The HIV-1/Human Protein Interaction Database is an online presentation of documented interactions between HIV-1 proteins, host cell proteins, other HIV-1 proteins or proteins from disease organisms associated with HIV or AIDS (50) . These data are maintained by the Division of Acquired Immunodeficiency Syndrome of the National Institute of Allergy and Infectious Diseases (NIAID) in collaboration with the Southern Research Institute and NCBI.

The BLAST programs (51-53) perform sequence-similarity searches against a variety of nucleotide and protein databases, returning a set of gapped alignments with links to full sequence records and related NCBI resources. The basic BLAST programs are also available as standalone command line programs and network clients at ftp.ncbi.nlm.nih. gov/blast/executables/LATEST/ (Table 2) .

NCBI provides an experimental Amazon Machine Image (AMI) for BLAST hosted at the Amazon Web Services (AWS) Marketplace. This AMI is preconfigured with the latest BLAST+ applications and includes a FUSE client that can download BLAST databases from NCBI as needed. Users can also upload their own custom databases. For tools that access NCBI BLAST programmatically, the AMI also supports the BLAST URL API, making an AWS instance a drop-in replacement for the NCBI BLAST website. More information is available at the BLAST help page as well as in an archived webinar about this AMI (www. ncbi.nlm.nih.gov/education/webinars/).

The default database for nucleotide BLAST searches (nr/nt) contains all RefSeq RNA records plus all GenBank sequences except for those from the EST, GSS, STS and HTG divisions. Another featured database is Human genomic plus transcript that contains human RefSeq transcript and genomic sequences arising from the NCBI annotation of the human genome. A similar database is available for mouse. Additional databases are also available and are described in links from the BLAST search form. Each of these databases can be limited to an arbitrary taxonomic node or those records satisfying any Entrez query.

For proteins the default database (nr) is a non-redundant set of all CDS translations from GenBank along with all sequences from RefSeq, UniProtKB/Swiss-Prot, PDB and the Protein Research Foundation (PRF). Subsets of this database are also available, such as the PDB or UniProtKB/Swiss-Prot sequences, along with separate databases for sequences from patents and environmental samples. Like the nucleotide databases, these collections can be limited by taxonomy or an arbitrary Entrez query.

Standard BLAST output formats include the default pairwise alignment, several query-anchored multiple sequence alignment formats, an easily-parsable Hit Table and a report that organizes the BLAST hits by taxonomy. A pairwise with identities mode better highlights differences between the query and a target sequence. A Tree View option for the Web BLAST service creates a dendrogram that clusters sequences according to their distances from the query sequence. Each alignment returned by BLAST is scored and assigned a measure of statistical significance called the Expectation Value (E-value). The alignments returned can be limited by an E-value threshold or range.

NCBI maintains Genomic BLAST services that mirror the design of the standard BLAST forms and allow users access to specialized databases for each particular genome. The default database contains the genomic sequence of an organism, but additional databases are provided depending on the available data and annotations. The default algorithm for Genomic BLAST is MegaBLAST (54), a faster version of standard nucleotide BLAST designed to find alignments between nearly identical sequences, typically from the same species. For rapid cross-species nucleotide queries, NCBI offers Discontiguous MegaBLAST, which uses a non-contiguous word match (55) as the nucleus for its alignments. Discontiguous MegaBLAST is far more rapid than a translated search such as blastx, yet maintains a competitive degree of sensitivity when comparing coding regions.

Primer-BLAST is a tool for designing and analyzing polymerase chain reaction (PCR) primers based on the existing program Primer3 (56) that designs PCR primers given a template DNA sequence. Primer-BLAST extends this functionality by running a BLAST search against a chosen database with the designed primers as queries, and then returns only those primer pairs specific to the desired target. If a user provides only one primer with the DNA template, the other primer will be designed and analyzed. If a user provides both primers and a template, the tool performs only the final BLAST analysis. If a user provides both primers but no template, primer-BLAST will display those templates that best match the primer pair. The available databases include the RefSeq mRNA collection, the BLAST nr database and genomic sets for one of twelve model organisms.

IgBLAST is a specialized BLAST tool that facilitates the analysis of immunoglobulin variable domain sequences and T-cell receptor sequences (57) . In addition to a standard BLAST analysis, IgBLAST reports the germline V, D and J gene matches to the query sequence, annotates immunoglobulin domains, reveals V(D)J junction details and indicates whether the rearrangement is in-frame or out-offrame. IgBLAST is available both as a web tool and as a stand-alone package.

The Constraint-based multiple protein Alignment Tool (COBALT) (58) is a multiple alignment algorithm for proteins that finds a collection of pair-wise constraints derived from both the NCBI CDD and the sequence similarity programs RPS-BLAST, BLASTp and PHI-BLAST. These pairwise constraints are then incorporated into a progressive multiple alignment. Links at the top of the COBALT report provide access to a phylogenetic tree view of the multiple alignment and allow users either to launch a modified search or download the alignment in several popular formats. 

PubChem (59,60) focuses on the chemical, structural and biological properties of small molecules, in particular their roles as diagnostic and therapeutic agents. A suite of three Entrez databases, PCSubstance, PCCompound and PCBioAssay, contain the structural and bioactivity data of the PubChem project. PubChem also provides a diverse set of three-dimensional (3D) conformers for 90% of the records in the PubChem Compound database.

The Biosystems database collects together molecules represented in Gene, Protein and PubChem that interact in a biological system, such as a biochemical pathway or disease. Currently Biosystems receives data from the Kyoto Encyclopedia of Genes and Genomes (KEGG) (61-63), Bio-Cyc (64), Reactome (65), the Pathway Interaction Database (66), WikiPathways (67, 68) and Gene Ontology (69) .

The resources described here include documentation, other explanatory material and references to collaborators and data sources on their respective websites. An alphabetical list of NCBI resources is available from a link above the category list on the left side of the NCBI home page. The NCBI Help Manual and the new second edition of the NCBI Handbook (www.ncbi.nlm.nih.gov/books/ NBK143764/), both available as links in the common page footer, describe the principal NCBI resources in detail. The NCBI Learn page (www.ncbi.nlm.nih.gov/home/learn. shtml) provides links to documentation, tutorials, webinars, courses and upcoming conference exhibits. A variety of video tutorials are available on the NCBI YouTube channel that can be accessed through links in the standard NCBI page footer. A user-support staff is available to answer questions at info@ncbi.nlm.nih.gov. Updates on NCBI resources and database enhancements are described on the NCBI News site (www.ncbi.nlm.nih.gov/news/), NCBI social media sites (FaceBook, Twitter, Google+ and LinkedIn), the 'NCBI Insights' blog, and the several mailing lists and RSS feeds that provide updates on services and databases. Links to these resources are in the NCBI page footer and on the NCBI News site.

Funding for open access charge: Intramural Research Program of the National Institutes of Health, National Library of Medicine.

Conflict of interest statement. None declared.

The NIH genetic testing registry: a new, centralized database of genetic tests to enable access to comprehensive information and improve transparency

Assembly: a resource for assembled genomes at NCBI

Database resources of the National Center for Biotechnology Information

Entrez: molecular biology database and retrieval system

PubMed Central--three years old and growing stronger

Medical Subject Headings in Medlars

ClinVar: public archive of relationships among sequence variation and human phenotype

ACMG recommendations for reporting of incidental findings in clinical exome and genome sequencing

New models of collaboration in genome-wide association studies: the Genetic Association Information Network

Allelic genes of blood group antigens: a source of human mutations and cSNPs documented in the Blood Group Antigen Gene Mutation Database

The sequencing-based typing tool of dbMHC: typing highly polymorphic gene sequences

NCBI Reference Sequences (RefSeq): current status, new features and genome annotation policy

The Sequence Read Archive: explosive growth of sequencing data

Clone DB: an integrated NCBI resource for clone-associated data

BioProject and BioSample databases at NCBI: facilitating capture and organization of metadata

The NCBI Taxonomy database

Type material in the NCBI Taxonomy Database

Toward richer metadata for microbial sequences: replacing strain-level NCBI taxonomy taxids with BioProject, BioSample and Assembly records

) dbSNP: the NCBI database of genetic variation

Diversity of human copy number variation and multicopy genes

Public data archives for genomic structural variation

Virus Variation Resource-recent updates and future directions

NCBI viral genomes resource

Virus variation resources at the National Center for Biotechnology Information: dengue virus

Visualization of large influenza virus sequence datasets using adaptively aggregated trees with sampling-based subscale representation

NCBI Epigenomics: a new public resource for exploring epigenomic data sets

Gene: a gene-centered information resource at NCBI

Clinical laboratory reports in molecular pathology

Current status and new features of the Consensus Coding Sequence database

NCBI GEO: archive for functional genomics data sets-update

Pieces of the puzzle: expressed sequence tags and the catalog of human genes

McKusick's Online Mendelian Inheritance in Man (OMIM)

The mouse genome database (MGD): new features facilitating a model system

The Zebrafish Information Network: the zebrafish model organism database

Gene Ontology annotations at SGD: new data sources and annotation methods

FlyBase: genomes by the dozen

UniProt Knowledgebase: a hub of integrated protein data

The worldwide Protein Data Bank (wwPDB): ensuring a single, uniform archive of PDB data

MMDB: 3D structures and macromolecular interactions

CDD: NCBI's conserved domain database

Cn3D: sequence and structure views for Entrez

Surprising similarities in structure comparison

Threading a database of protein cores

CDD: specific functional annotation with the Conserved Domain Database

SMART 5: domains in the context of genomes and networks

Pfam: clans, web tools and services

The TIGRFAMs database of protein families

The National Center for Biotechnology Information's Protein Clusters Database

Human immunodeficiency virus type 1, human protein interaction database at NCBI

Basic local alignment search tool

Gapped BLAST and PSI-BLAST: a new generation of protein database search programs

BLAST: a more efficient report with usability improvements

A greedy algorithm for aligning DNA sequences

PatternHunter: faster and more sensitive homology search

Primer3 on the WWW for general users and for biologist programmers

IgBLAST: an immunoglobulin variable domain sequence analysis tool

COBALT: constraint-based alignment tool for multiple protein sequences

PubChem: a public information system for analyzing bioactivities of small molecules

PubChem's BioAssay Database

KEGG for linking genomes to life and the environment

KEGG: kyoto encyclopedia of genes and genomes

From genomics to chemical genomics: new developments in KEGG

EcoCyc: a comprehensive view of Escherichia coli biology

Reactome knowledgebase of human biological pathways and processes

PID: the Pathway Interaction Database

Mining biological pathways using WikiPathways web services

WikiPathways: pathway editing for the people

Gene ontology: tool for the unification of biology. The Gene Ontology Consortium