Europe PMC

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

The DisGeNET knowledge platform for disease genomics: 2019 update.

Author information

Affiliations

  • 1. Research Programme on Biomedical Informatics (GRIB), Hospital del Mar Medical Research Institute (IMIM), Department of Experimental and Health Sciences, Pompeu Fabra University (UPF), Barcelona, Spain.
      Authors
    • Piñero J 1
    • Ramírez-Anguita JM 1
    • Saüch-Pitarch J 1
    • Ronzano F 1
    • Centeno E 1
    • Sanz F 1
    • Furlong LI 1
    (7 authors)

ORCIDs linked to this article

Nucleic Acids Research, 01 Jan 2020, 48(D1):D845-D855
https://doi.org/10.1093/nar/gkz1021 PMID: 31680165 PMCID: PMC7145631

This article is in the Europe PMC Open access subset. Refer to the copyright information in the article for licensing details.
Free full text in Europe PMC

Abstract 

One of the most pressing challenges in genomic medicine is to understand the role played by genetic variation in health and disease. Thanks to the exploration of genomic variants at large scale, hundreds of thousands of disease-associated loci have been uncovered. However, the identification of variants of clinical relevance is a significant challenge that requires comprehensive interrogation of previous knowledge and linkage to new experimental results. To assist in this complex task, we created DisGeNET (http://www.disgenet.org/), a knowledge management platform integrating and standardizing data about disease associated genes and variants from multiple sources, including the scientific literature. DisGeNET covers the full spectrum of human diseases as well as normal and abnormal traits. The current release covers more than 24 000 diseases and traits, 17 000 genes and 117 000 genomic variants. The latest developments of DisGeNET include new sources of data, novel data attributes and prioritization metrics, a redesigned web interface and recently launched APIs. Thanks to the data standardization, the combination of expert curated information with data automatically mined from the scientific literature, and a suite of tools for accessing its publicly available data, DisGeNET is an interoperable resource supporting a variety of applications in genomic medicine and drug R&D.

Free full text 

Logo of narLink to Publisher's site
Nucleic Acids Res. 2020 Jan 8; 48(D1): D845–D855.
Published online 2019 Nov 4. https://doi.org/10.1093/nar/gkz1021
PMCID: PMC7145631
PMID: 31680165

The DisGeNET knowledge platform for disease genomics: 2019 update

Janet Piñero, Juan Manuel Ramírez-Anguita, Josep Saüch-Pitarch, Francesco Ronzano, Emilio Centeno, Ferran Sanz, and Laura I Furlong
Author information Article notes Copyright and License information Disclaimer

Abstract

One of the most pressing challenges in genomic medicine is to understand the role played by genetic variation in health and disease. Thanks to the exploration of genomic variants at large scale, hundreds of thousands of disease-associated loci have been uncovered. However, the identification of variants of clinical relevance is a significant challenge that requires comprehensive interrogation of previous knowledge and linkage to new experimental results. To assist in this complex task, we created DisGeNET (http://www.disgenet.org/), a knowledge management platform integrating and standardizing data about disease associated genes and variants from multiple sources, including the scientific literature. DisGeNET covers the full spectrum of human diseases as well as normal and abnormal traits. The current release covers more than 24 000 diseases and traits, 17 000 genes and 117 000 genomic variants. The latest developments of DisGeNET include new sources of data, novel data attributes and prioritization metrics, a redesigned web interface and recently launched APIs. Thanks to the data standardization, the combination of expert curated information with data automatically mined from the scientific literature, and a suite of tools for accessing its publicly available data, DisGeNET is an interoperable resource supporting a variety of applications in genomic medicine and drug R&D.

INTRODUCTION

Modern genome sequencing technologies are fostering the integration of genomics into clinical practice. The exploration of human variation at large scale by genome sequencing or SNP array genotyping are enabling the identification of disease-associated variants for a wide range of diseases and conditions. Nevertheless, the interpretation of the results of genomic analysis and the identification of variant of clinical relevance remain a significant challenge (1). Variant assessment still involves manual exploration of multiple sources of data, which requires a significant amount of time and experts in the domain. In this context, new bioinformatic tools and resources that enable the automation of every possible step in this process are crucial. In this regard, resources such as ClinVar (2), ClinGen (3), the Genomics England PanelApp (https://panelapp.genomicsengland.co.uk/), Orphanet (4) and OMIM (5), among others, have demonstrated their utility to support variant interpretation. Here, we present a new release of DisGeNET, a knowledge management platform that houses one of the most exhaustive and publicly available catalogues of genes and genomic variants associated with human diseases. Originally implemented in 2010 as a Cytoscape plugin (6), during the last years DisGeNET has evolved into different formats and tools (7–10), and it now undergoes its sixth release as a knowledge management platform aimed at supporting different application scenarios and users.

The DisGeNET database contents

The core concepts in the DisGeNET database structure (Figure ​(Figure1A)1A) are the Gene–Disease Association (GDA) and the Variant–Disease Association (VDA), that are collated from different data sources (Figure ​(Figure2).2). The integration of these diverse sources of data is enabled by proper standardization of genes, variants, diseases (diseases, symptoms and traits) and associations using community-driven ontologies and controlled vocabularies, as well as ontologies developed ad hoc (e.g. the DisGeNET association type ontology). Of note, the provenance of the information is provided in several ways: (a) as the field ‘original database’ that indicates where the data was taken from (e.g. ClinVar or UniProt (11)), (b) with the number of articles that support the association and the NCBI PMIDs of these publications and (c) with a text excerpt from the article that expresses the evidence for the association. GDAs and VDAs are further annotated with in-house and external attributes easing data analysis, exploration and prioritization. For the attributes incorporated from external resources the provenance is also provided.

An external file that holds a picture, illustration, etc. Object name is gkz1021fig1.jpg

The DisGeNET platform. (A) Simplified DisGeNET database schema. (B) Tools to access DisGeNET data.

An external file that holds a picture, illustration, etc. Object name is gkz1021fig2.jpg

Data sources and data types in DisGeNET. For Gene–Disease Associations (GDAs) the data sources are classified as Curated, Animal models, Inferred and Literature. For Variant-Disease Associations (VDAs) the data sources are classified as Curated and Literature. The data sources in white are developed in-house, while the others are third-party resources.

New data sources and data types

The current release (v6.0) of the database contains 628 685 gene-disease associations (GDAs), involving 17 549 genes and 24 166 diseases, and 210 498 variant-disease associations (VDAs), including 117 337 variants and 10 358 diseases (see details in Table ​Table1).1). Note that the term ‘disease’ refers to a wide range of phenotypes relevant in human genomics: actual diseases, disease symptoms and abnormal phenotypes that are observed as disease manifestations, as well as normal traits and phenotypes that are currently explored in large scale Genome Wide Association studies (GWAs) (see section New data attributes and prioritization metrics for more details on disease standardization and annotation). The GDAs and VDAs integrated in the DisGeNET database originate from over a dozen repositories, each one with a different focus, for example, databases that annotate clinically relevant variants (ClinVar) or genes (ClinGen, Genomics England PanelApp, among others), or specialized in certain disease classes (e.g. Orphanet for rare diseases) or compiling information from animal models of disease (e.g. MGD (12) and RGD (13)) (Figure ​(Figure2).2). In addition to the original source of information for the VDAs and GDAs, DisGeNET provides a classification for the database sources: for the gene-disease associations (GDAs), the information is classified as Curated, Animal Models, Literature and a new category, Inferred (Figure ​(Figure22 and Table ​Table3).3). In the case of variant-disease associations (VDAs), the data is classified into Curated and Literature. For more details about the data content in DisGeNET 6.0, see Tables ​Tables11 and 2.

Table 1.

Distribution of genes, diseases and GDAs by source

SourceGenesDiseasesAssocs
CGI3412001650
CLINGEN274205518
GEN. ENGLAND33261147897
CTD_human7919825162 794
ORPHANET349635206850
PSYGENET15301093656
UNIPROT373045426798
CURATED 9413 10 370 81 746
HPO36887502134 890
CLINVAR3848630710 695
GWASDB39483218253
GWASCAT476765314 182
INFERRED 8700 13 176 163 626
CTD_mouse71298474
CTD_rat222646
MGD163721114711
RGD15856816364
ANIMAL MODELS 2795 2789 11 517
LHGDN5938180031431
BEFREE15 14712 219401 440
LITERATURE 15 283 12 418 415 583
ALL 17 549 24 166 628 685

Table 2.

Distribution of variants, diseases and VDAs by source

SourceVariantsDiseasesAssocs
CLINVAR50 141644367 978
GWASDB32 16238646 468
GWASCAT20 48672532 950
UNIPROT20 148424635 217
CURATED 104 653 7954 165 354
BEFREE19 407422848 998
LITERATURE 19 407 4228 48 998
ALL117 33710 358210 498

Table 3.

Classification of the data sources in DisGeNET

Source typeGDAsVDAs
CuratedUniProtUniProt
CTDClinVar
OrphanetGWAS Catalog
ClinGen*GWAS DB*
Genomics England*
CGI*
PsyGeNET
Animal modelsRGDNA
MGD
CTD
InferredHPONA
ClinVar
GWAS Catalog
GWAS DB*
LiteratureBeFreeBeFree
LHGDN

*New source with respect to the previous release 5.0. NA: not available.

Mining disease-associated genes and variants from the literature

A distinctive feature of DisGeNET is its unique collection of GDAs and VDAs extracted by text mining the scientific literature (14,15). DisGeNET contains a corpus of 400K publications with information about GDAs and VDAs. Sixty percent of the GDAs included in DisGeNET have been extracted from the scientific literature by text mining and are not reported in any of the curated resources integrated in DisGeNET. Due to the current challenge to manually identify, curate and properly store phenotype–genotype information as structured data, it is important to have means to extract this information from the literature in an automatic and exhaustive manner to keep the pace of the most recent scientific findings. The importance of collecting this information is particularly evident in the clinical genomics area, where there is a pressing need to identify all the knowledge, including the most recent one, on disease association for sequence variants identified in the genome of patients. An example of the insights that this expanded information can potentially provide over current authoritative resources and gene panels databases is illustrated in section ‘Expanding information for rare diseases’. Our text mining tools leverage on controlled vocabularies and ontologies to properly identify and standardize the entities and relationships found in the literature, and they exploit linguistic and semantic textual features to identify genotype-phenotype relationships. On the other hand, it is noteworthy that 78% of the GDAs and 91% of the VDAs reported in DisGeNET are supported by at least one bibliographic reference. In addition, to help the user in navigating literature-derived data, for each publication we provide an exemplary sentence or text excerpt that expresses the association under study (see Figure ​Figure7C7C for an example).

An external file that holds a picture, illustration, etc. Object name is gkz1021fig7.jpg

Analysis of GWAs results with DisGeNET. (A) 61 out of 143 variants identified by a recent GWAs of Type 2 Diabetes Mellitus (T2D) (29) are reported in DisGeNET as associated to cardiometabolic diseases and traits, and 47 variants are annotated to T2D. (B) Top-scoring variants in DisGeNET from those found in the study (29). DisGeNET provides additional information such as the consequence type of the variant according to VEP, allele frequencies from the gnomAD database, DisGeNET score, number of supporting publications with linkouts to MEDLINE, to name a few attributes. (C) Network of diseases and phenotypes associated with variant rs7903146 annotated by curated databases, created with the DisGeNET Cytoscape App. Examples of text excerpts extracted by text mining from publications supporting the association are shown.

Disease–disease associations

In this DisGeNET release we present a new dataset, the DisGeNET disease-disease associations (DDAs), which can be used to explore similarities between pairs of diseases or between diseases and traits based on shared genes and variants. The analysis of DDAs can support a variety of applications, such as the study of disease comorbidities as well as finding genomic similarities among different disease diagnosis. The DDAs are obtained by connecting two diseases if they share at least one gene or one variant in a particular source database (Figure ​(Figure3A).3A). The fraction of shared genes (or variants) between two diseases is assessed by the Jaccard Index (JI), and a P-value obtained by permutation testing (for more details see http://www.disgenet.org/dbinfo). The DDAs dataset contains more than 11 × 106 pairs of diseases sharing at least one gene (P-value < 0.05) and over 200 000 pairs of diseases sharing at least one variant (P-value < 0.05). The DDAs dataset can be explored via the web interface, where the user can search by disease or database source and apply different filters such as JI value, minimum number of shared genes (or variants), P-value threshold, disease class, among others. This new dataset is also available via the DisGENET REST-API, the disgenet2r package and the Cytoscape App (Figure ​(Figure3B3B).

An external file that holds a picture, illustration, etc. Object name is gkz1021fig3.jpg

Disease-Disease associations in DisGeNET. (A) Two diseases are connected if they share at least one gene or one variant in the GDA or the VDA dataset, respectively. A Jaccard Index with its associated P-value are provided for each association to rank and filter the Disease-Disease association results. For more details see http://www.disgenet.org/dbinfo. (B) The Disease-Disease association network of Type 2 Diabetes Mellitus (T2D, CUI: C0011860). The network shows the diseases associated to T2D through common variants from DisGeNET curated databases with a P-value ≤ 0.05.

New data attributes and prioritization metrics

Diseases, genes and variants in DisGeNET are annotated with attributes that cover a wide variety of biomedical resources: diseases are coded using UMLS® (16) concept unique identifiers (CUIs), and annotated with the UMLS® semantic type, the MeSH class, and the top level concepts from the Human Disease Ontology (17) and the Human Phenotype Ontology (18). Genes are referred by their NCBI identifiers, and are annotated with the official gene symbol, the UniProt accession, the protein class, and with their value of pLI (probability of being loss-of-function intolerant), a gene constraint metric from the gnomAD consortium (19). The genomic variants are identified using the dbSNP identifier and annotated with their reference and alternative alleles, and their genomic coordinates (corresponding to NCBI dbSNP Human Build 151, and Assembly GRCh38). Additionally, the variants are classified according to their most severe consequence type assigned by the Variant Effect Predictor (20) based on canonical gene transcripts. Most DisGeNET variants are missense (28%), followed by intronic (26%), frameshift and intergenic (both 11%) (Figure ​(Figure4).4). In this new release of DisGeNET, variants are also annotated with the allelic frequency in genomes and exomes according to data from the gnomAD consortium.

An external file that holds a picture, illustration, etc. Object name is gkz1021fig4.jpg

Distribution of most severe consequence types in DisGeNET variants. Consequence types are obtained from the Variant Effect Predictor (ENSEMBL).

Additionally, the platform includes a series of in-house developed metrics and attributes to facilitate ranking and filtering the information. Each phenotypic concept is classified according to a DisGeNET type as disease (such as Crohn disease, schizophrenia, Alzheimer's disease, etc.), phenotype (such as depressive symptoms, blood pressure, body mass index, neutrophil count, etc.) or group (such as cardiovascular diseases, neoplasms, etc.). This classification is based on the UMLS semantic types and expert curation. Sixty-six percent of DisGeNET CUIs are classified as diseases, 4% are classified as groups, and 30% as phenotypes. This release of DisGeNET includes a larger number of phenotypes because in this class are included traits, measurements and laboratory test results that are collected mainly by the GWAS catalog (21) and GWASdb (22).

While 20% of the genes in the Curated DisGeNET subset (12% in the whole DisGeNET database) are annotated to a single disease or phenotype concept, the remaining genes are annotated to more than one disease or phenotype, with exceptional cases of clinically relevant genes annotated to over hundreds or thousands of concepts, such as TP53, TNF, PTEN and MTHFR (Figure ​(Figure5A).5A). A similar behavior is observed for the variants, although the fraction of variants annotated to a single concept is higher than for genes (over 60%, Figure ​Figure5B).5B). In this regard, we define the Disease Specificity Index (DSI) and the Disease Pleiotropy Index (DPI) to reflect the different behaviour of genes and variants with respect to the number of associated diseases (see http://www.disgenet.org/dbinfo#DPI and http://www.disgenet.org/dbinfo#DSI for more details). Both metrics are aimed at indicating how specific is a gene or variant with respect to the associated diseases. A value of the DSI close to one means that the gene or variant is disease-specific, while a value close to zero indicates that the gene or variant is disease-promiscuous. The DPI considers if the diseases associated with the gene (or variant) are similar among them and belong to the same disease class (e.g. Cardiovascular Diseases) or belong to different disease classes. In this case, disease–promiscuous genes or variants generate values of DPI close to one.

An external file that holds a picture, illustration, etc. Object name is gkz1021fig5.jpg

Distribution of number of associated diseases per gene (panel A) and variant (panel B) in the DisGeNET Curated subset and in the whole database (ALL). Note that genes or variants associated to a single UMLS concept have a DSI equal to one, and a DPI close to zero, while genes or variants associated to several UMLS concepts have higher DPI, and lower DSI.

DisGeNET provides several attributes and metrics that allow the user to evaluate the relevance of the gene and variant associations, which is especially helpful in the case of diseases with a large number of associated genes or variants (for instance Schizophrenia has >1000 genes and variants in Curated sources, and over 1700 in the whole database). The DisGeNET association type provides a semantic classification of the biology of the association. The Evidence Level, only available for GDAs coming from ClinGen and Genomics England PanelApp, classifies the association according to the available evidence based on expert assessment in these databases (23). The number of supporting publications indicates how well studied is the association, along with its temporal span (year of first and last publication recorded in DisGeNET, see Figure ​Figure6A6A for an example). This last feature can also be used to distinguish novel associations from those well described associations having a large number of publications and to identify new trends in the field of disease genetics and genomics. The DisGeNET score is an in-house developed metric that reflects how well established is a particular association based on current knowledge. The DisGeNET score gives the highest value to associations that are reported by several databases, in particular to those reported by expert curated resources, and with a large number of supporting publications. More details on how the score is calculated are provided in the online documentation (http://www.disgenet.org/dbinfo#GDAScore). Finally, both GDAs and VDAs are annotated with the Evidence Index (EI), which indicates the existence of contradictory results in the publications supporting the associations. This index is computed for the associations coming from BeFree and PsyGeNET (24), which identify the publications reporting a negative finding on a particular VDA or GDA. Note that only in the case of PsyGeNET the information used to compute the EI has been validated by experts. An EI equal to one indicates that all the publications support the GDA or the VDA, while an EI smaller than one indicates that there are publications that assert that there is no association between the gene/variants and the disease.

An external file that holds a picture, illustration, etc. Object name is gkz1021fig6.jpg

DisGeNET provides comprehensive information on genes and variants for rare diseases. (A) Highest scoring genes associated to Duchenne Muscular Dystrophy in DisGeNET. In blue background, the genes annotated by CURATED resources. (B) DisGeNET annotates 350 DMD variants to Duchenne muscular dystrophy, being most of them stop-gained variants. (C) The DMD gene is associated to a large number of diseases and phenotypes belonging to different disease classes. (D) Pathways associated with Duchenne Muscular Dystrophy obtained by a federated query interrogating DisGeNET and WikiPathways.

DisGeNET tools

DisGeNET is available via a suite of tools (Figure ​(Figure1B)1B) described in more detail in the next section.

The DisGeNET web interface

DisGeNET 6.0 benefits from a completely new web interface aimed at improving the user experience. The Search functionality of the web interface supports searches by single disease, gene, or variant, as well as lists of these entities. The Browse functionality allows exploring DisGeNET by the source databases, for example CURATED. The results of the searches are organized in tables providing different views of the information: summaries of GDAs and VDAs, evidence supporting the associations, or views focused on diseases, genes or variants. The results of the browsing or the initial searches can be furthered filtered and prioritized using a collection of flexible filters that can be used alone or in combination. For example, diseases might be filtered by UMLS semantic type, and/or by MeSH disease class. Genes might be filtered by protein class, or by values of the DPI, DSI or the pLI. The variants might be filtered by their consequence type, and their allelic frequency in exome and genome data. The results of the searches can be downloaded as tabulated or Microsoft Excel files, or shared by using a URL.

The DisGeNET REST API

DisGeNET 6.0 features a new REST application programming interface (API) that enables programmatic retrieval of the information contained in the platform. The base URL for the DisGeNET REST API is http://www.disgenet.org/api/. The services in the API allow to retrieve GDAs, VDAs, DDAs and attributes of genes, diseases, and variants in different formats (json, xml and tsv). Additionally, the API includes a service that provides mappings between different disease identifiers from a variety of sources such as UMLS, MeSH, OMIM, HPO, DO, MONDO, NCI and ICD-9.

The DisGeNET-RDF dataset

The DisGeNET-RDF Linked Dataset is an alternative way to access the DisGeNET data and enables the integration and joint querying of the DisGeNET data with other databases available as Linked Open Data (https://lod-cloud.net/). DisGeNET-RDF (8) encodes DisGeNET data using the Resource Description Framework (RDF) (http://www.w3.org/RDF/), which is a core part of Semantic Web standards. The main components of the DisGeNET-RDF dataset are the GDA and VDA datasets, metadata description of the RDF dataset (VoID description), and linkouts to other Linked Datasets. The RDF representation of DisGeNET (v6.0.0) contains 62,359,775 triples serialized in Turtle syntax that annotate the 628,685 GDAs and 210,498 VDAs contained in DisGeNET 6.0. Entities and properties are semantically defined using standard ontologies such as the National Cancer Institute thesaurus (NCIt), and resources identified by using de-referenceable IRIs. GDAs are integrated using the DisGeNET Association Type Ontology and they are semantically harmonized using SIO classes (25).

By relying on the web-based data representation and integration framework known as Linked Open Data, that constitutes the backbone of the Semantic Web, DisGeNET-RDF enables sharing and integration of the DisGeNET data with external resources such as databases on gene expression, drugs and chemicals, proteins, biological pathways and systems biology models. Through the SPARQL endpoint, query federation to interrogate DisGeNET in combination with these LOD resources is possible. Examples of research questions that can be addressed using DisGeNET-RDF are provided at disgenet.org/rdf. The DisGeNET-RDF API has been recently selected as one of the 10 interoperability resources recommended by ELIXIR (https://elixir-europe.org/platforms/interoperability/rirs), the European organisation that brings together bioinformatics resources in life sciences, to facilitate interoperability and reusability of life science data and support the principles of FAIR data management.

The DisGeNET Cytoscape App

The DisGeNET Cytoscape App contains a set of functions to query, analyze, and visualize DisGeNET data from a network biology perspective. The GDAs, VDAs and DDAs can be represented, queried and filtered as bipartite and monopartite networks. Note that VDAs are a new feature in this release of the DisGeNET App. The new version of the App includes functions to query DisGeNET data for specific diseases, genes, and variants, and their combinations, and for filtering the information by source, the DisGeNET score, DisGeNET association type, Evidence Index and disease class. Note that VDAs, the DisGeNET score and Evidence Index are a new feature in this release of the DisGeNET App. Another novelty is a function for the annotation of entities from foreign networks with DisGeNET data, such as protein, gene or variants generated by other Cytoscape Apps or networks uploaded by the user. Finally, the App features a new automation module that includes a set of functions to programmatically execute different functionalities using popular programming languages such as R and Python. In summary, with its newly implemented capabilities, the DisGeNET Cytoscape App provides the biomedical community a tool that enables a systems-level analysis of human diseases in an easy and automatic way.

The disgenet2r package

The disgenet2r package has been updated for the new database framework. The package allows to easily interact with both the REST API and the SPARQL API and provides a series of plots for the visualization of DisGeNET data which include networks, heatmaps and Venn diagrams. Additionally, a suite of functions has been incorporated to interrogate DisGeNET-RDF and perform federated queries such as the retrieval of all the pathways for a particular disease, or the search of the disease-associated proteins that are also drug targets.

Data downloads

The DisGeNET database is made available under the Attribution-NonCommercial-ShareAlike 4.0 International License. The complete database and different subsets of data are available for download as tabulated files (http://www.disgenet.org/downloads) as well as the results of specific searches carried out through the web interface.

Application examples of DisGeNET

Expanding information for rare diseases

DisGeNET integrates information from databases with a special focus on rare diseases (such as Orphanet) with data coming from other resources, significantly expanding the number of genes, variants and publications annotated to those diseases. An illustrative example of this expansion is presented here with Duchenne muscular dystrophy, a neuromuscular disease characterized by rapidly progressive muscle weakness and wasting. This severe, inherited X-linked recessive disease has no current treatment beyond symptoms management. Muscle damage is caused by absence of the sarcolemmal protein dystrophin as a result of DMD gene mutations.

Two genes are annotated to the disease in Orphanet (DMD, LTB4, https://www.orpha.net/consor/cgi-bin/Disease_Genes_Simple.php?lng=EN&LnkId=13913&Typ=Pat&diseaseType=Gen&from=rightMenu) and one in OMIM (DMD, https://www.omim.org/entry/310200?search=duchenne%20muscular%20dystrophy&highlight=duchenne%20dystrophy%20muscular). Contrastingly, DisGeNET provides 189 genes, 6 of them from CURATED resources (in blue in Figure ​Figure6A),6A), as well as 353 variants (most of them from ClinVar, an expert curated resource on clinical genomics). The top 15 genes ranked by DisGeNET score are shown in Figure ​Figure6A.6A. The DMD gene has the highest score, while the other genes have lower score mainly because they are reported in a lower number of databases and/or in fewer publications (column NPMIDS in the table shown in Figure ​Figure6A).6A). DisGeNET annotates 350 DMD variants to Duchenne muscular dystrophy, being most of them stop gained variants (Figure ​(Figure6B)6B) located throughout the protein coding sequence. An interesting example among the list of genes associated with Duchenne muscular dystrophy is the gene UTRN encoding the utrophin protein whose increased levels have previously been shown to compensate in part for the loss of dystrophin (26) and proposed to play a role as disease modifier. The role of utrophin in the disease has been studied since 1991 (44 publications listed in DisGeNET) and the effect of its expression is currently being investigated by genome editing technologies (27). DisGeNET also indicates that the DMD gene is associated to a large number of diseases and phenotypes (almost 300 UMLS CUIs), as reflected by its low DSI. It is associated with different types of muscular dystrophies (Duchenne and Becker Muscular Dystrophy), cardiovascular diseases (Dilated and Familial Cardiomyopathy), and mental diseases (Impaired Cognition and Intellectual Disability), among others (Figure ​(Figure6C).6C). By performing federated queries to jointly interrogate DisGeNET-RDF and WikiPathways (28), it is possible to identify the pathways associated with the disease (Figure ​(Figure6D).6D). Of note, pathways related to cardiomyopathy (Viral acute myocarditis, Arrhythmogenic Right Ventricular Cardiomyopathy, and Striated Muscle Contraction), spinal cord injury, and several signalling pathways, all processes related to the disease pathophysiology, concentrate the largest number of genes. In summary, DisGeNET significantly expands information on genes and variants associated to rare diseases, which can be exploited for development of clinical genomics pipelines in this area and supporting research and development of new therapies.

Analysis of NGS and GWAs data for complex diseases

DisGeNET can also be used for the analysis and interpretation of genomic data from studies on complex diseases and traits. A recent meta-analysis of GWAs identified 143 risk variants for type 2 diabetes (T2D) through the study of 62 000 T2D cases and 596 000 controls of European ancestry (29). DisGeNET, as a database that aggregates available knowledge on disease relevance of genomic variants, can aid in the analysis of the GWAs signals, in particular to identify those variants already reported to be associated with the disease or trait under study. From the list of 143 variants identified by Xue and colleagues, 61 are reported in DisGeNET as associated to cardiometabolic diseases and traits (Figure ​(Figure7A),7A), and from them, 47 variants are annotated to T2D. Notably, two of these variants (rs10401969, rs7674212) are reported as novel independent risk loci not previously associated to T2D by Xue and colleagues (see Table ​Table11 in ref. (29)), although there are publications that report their association to diabetes and metabolic traits dating from 2011 in DisGeNET (http://www.disgenet.org/browser/2/1/0/rs7674212::rs10401969/). For the disease associated variants, DisGeNET provides genomic information such as the consequence type of the variant according to VEP, allele frequencies from the gnomAD databases, along with detailed information on disease and phenotype annotation including the DisGeNET score, Evidence Index, number of supporting publications with linkouts to MEDLINE abstracts and the text excerpt asserting the VDA. In addition, for each VDA, the first and last year of reference publications are provided (Figure ​(Figure7B).7B). The DSI and DPI can be used to select variants according to their specificity for the disease (Figure ​(Figure7B).7B). Figure ​Figure7C7C shows the diseases associated with variant rs7903146, an intronic variant in the gene TCF7L2. The association with T2D has a high DisGeNET score (0.9) and is supported by 138 publications dating from 2006. Notably, the T allele of rs7903146 variants confers the strongest risk of T2D known to date in Caucasians (P < 10−347) (29), and is a common variant with AF of 0.26 in gnomAD. Moreover, the literature on VDAs captured in DisGeNET can provide insights on putative mechanisms by which the variant confer risk to the disease (by modification of the effect of incretins on insulin secretion (30–32)). In summary, DisGeNET, as a publicly available knowledge management tool and comprehensive database enables efficient analysis and interpretation of GWAs.

CONCLUSION

The DisGeNET platform is designed to allow easy exploration and analysis of the genetic underpinnings of the full spectrum of human diseases: Mendelian, rare and complex, as well as symptoms, signs and other phenotypic manifestations of diseases. The platform contains data from the most popular repositories in the field that have been enriched and expanded with information extracted from the scientific literature using state-of-the-art text mining tools and mostly not reported in any other repository.

The data in DisGeNET is harmonized and standardized by controlled vocabularies and ontologies following the FAIR principles, which enables an easy link with other biomedical resources. This interoperability is particularly facilitated by providing an RDF version of the DisGeNET database. In addition, the information is enriched with a series of in-house developed metrics and external attributes facilitating its interpretation and analysis, both manual and automatic. For a better assessment of the genotype-phenotype associations, DisGeNET provides information about their original source, links to the publications that support the associations, as well as a representative sentence of each publication. In addition to the possibility of downloading data in various formats, DisGeNET offers a series of bioinformatics tools to facilitate access and analysis of data: a web interface, a Cytoscape App, an R package and different APIs (SPARQL endpoint, Rest API, Cytoscape Automation).

Since its first release, DisGeNET has become an established and mature resource, broadly used by the biomedical community, enabling a wide variety of applications in the field of drug R&D, disease genomics and for the development of bioinformatic tools and databases.

FUNDING

ISCIII-FEDER [PI13/00082, PI17/00230, CPII16/00026]; IMI-JU resources of which are composed of financial contribution from the EU-FP7 [FP7/2007–2013] and EFPIA companies in kind contribution [116030 to TransQST, 777365 to eTRANSAFE], and the EU H2020 Programme 2014–2020 [676559 to Elixir-Excelerate]; Agència de Gestió d’Ajuts Universitaris i de Recerca Generalitat de Catalunya [2017SGR00519]. The Research Programme on Biomedical Informatics (GRIB) is a member of the Spanish National Bioinformatics Institute (INB), funded by ISCIII and FEDER (PRB2-ISCIII [PT13/0001/0023, of the PE I+D+i 2013–2016]). The DCEXS is a ‘Unidad de Excelencia María de Maeztu’, funded by the MINECO [MDM-2014-0370]. Funding for open access charge: Agència de Gestió d’Ajuts Universitaris i de Recerca Generalitat de Catalunya [2017SGR00519].

Conflict of interest statement. None declared.

REFERENCES

1. Eilbeck K., Quinlan A., Yandell M.. Settling the score: variant prioritization and Mendelian disease. Nat. Rev. Genet. 2017; 18:599–612. [Europe PMC free article] [Abstract] [Google Scholar]
2. Landrum M.J., Lee J.M., Benson M., Brown G.R., Chao C., Chitipiralla S., Gu B., Hart J., Hoffman D., Jang W. et al. .. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 2018; 46:D1062–D1067. [Europe PMC free article] [Abstract] [Google Scholar]
3. Rehm H.L., Berg J.S., Brooks L.D., Bustamante C.D., Evans J.P., Landrum M.J., Ledbetter D.H., Maglott D.R., Martin C.L., Nussbaum R.L. et al. .. ClinGen — The Clinical Genome Resource. N. Engl. J. Med. 2015; 372:2235–2242. [Europe PMC free article] [Abstract] [Google Scholar]
4. Rath A., Olry A., Dhombres F., Brandt M.M., Urbero B., Ayme S.. Representation of rare diseases in health information systems: the orphanet approach to serve a wide range of end users. Hum. Mutat. 2012; 33:803–808. [Abstract] [Google Scholar]
5. Amberger J.S., Bocchini C.A., Scott A.F., Hamosh A.. OMIM.org: leveraging knowledge across phenotype–gene relationships. Nucleic Acids Res. 2019; 47:D1038–D1043. [Europe PMC free article] [Abstract] [Google Scholar]
6. Bauer-Mehren A., Rautschka M., Sanz F., Furlong L.I.. DisGeNET: a Cytoscape plugin to visualize, integrate, search and analyze gene-disease networks. Bioinformatics. 2010; 26:2924–2926. [Abstract] [Google Scholar]
7. Piñero J., Queralt-Rosinach N., Bravo À., Deu-Pons J., Bauer-Mehren A., Baron M., Sanz F., Furlong L.I.. DisGeNET: a discovery platform for the dynamical exploration of human diseases and their genes. Database. 2015; 2015:bav028. [Europe PMC free article] [Abstract] [Google Scholar]
8. Queralt-Rosinach N., Piñero J., Bravo À., Sanz F., Furlong L.I.. DisGeNET-RDF: harnessing the innovative power of the Semantic Web to explore the genetic basis of diseases. Bioinformatics. 2016; 32:2236–2238. [Europe PMC free article] [Abstract] [Google Scholar]
9. Queralt-Rosinach N., Kuhn T., Chichester C., Dumontier M., Sanz F., Furlong L.I.. Publishing DisGeNET as nanopublications. Semant. Web. 2016; 7:519–528. [Google Scholar]
10. Piñero J., Bravo A., Queralt-Rosinach N., Gutiérrez-Sacristán A., Deu-Pons J., Centeno E., García-García J., Sanz F., Furlong L.I.. DisGeNET: A comprehensive platform integrating information on human disease-associated genes and variants. Nucleic Acids Res. 2017; 45:D833–D839. [Europe PMC free article] [Abstract] [Google Scholar]
11. UniProt Consortium UniProt: a worldwide hub of protein knowledge. Nucleic Acids Res. 2019; 47:D506–D515. [Europe PMC free article] [Abstract] [Google Scholar]
12. Bult C.J., Blake J.A., Smith C.L., Kadin J.A., Richardson J.E., Anagnostopoulos A., Asabor R., Baldarelli R.M., Beal J.S., Bello S.M. et al. .. Mouse Genome Database (MGD) 2019. Nucleic Acids Res. 2019; 47:D801–D806. [Europe PMC free article] [Abstract] [Google Scholar]
13. Shimoyama M., De Pons J., Hayman G.T., Laulederkind S.J.F., Liu W., Nigam R., Petri V., Smith J.R., Tutaj M., Wang S.-J. et al. .. The Rat Genome Database 2015: genomic, phenotypic and environmental variations and disease. Nucleic Acids Res. 2015; 43:D743–D750. [Europe PMC free article] [Abstract] [Google Scholar]
14. Bravo A., Piñero J., Queralt-Rosinach N., Rautschka M., Furlong L.I.. Extraction of relations between genes and diseases from text and large-scale data analysis: Implications for translational research. BMC Bioinformatics. 2015; 16:55. [Europe PMC free article] [Abstract] [Google Scholar]
15. Thomas P., Rocktäschel T., Hakenberg J., Lichtblau Y., Leser U.. SETH detects and normalizes genetic variants in text. Bioinformatics. 2016; 32:2883–2885. [Abstract] [Google Scholar]
16. Bodenreider O. The Unified Medical Language System (UMLS): integrating biomedical terminology. Nucleic Acids Res. 2004; 32:267D–270. [Europe PMC free article] [Abstract] [Google Scholar]
17. Schriml L.M., Mitraka E., Munro J., Tauber B., Schor M., Nickle L., Felix V., Jeng L., Bearer C., Lichenstein R. et al. .. Human Disease Ontology 2018 update: classification, content and workflow expansion. Nucleic Acids Res. 2019; 47:D955–D962. [Europe PMC free article] [Abstract] [Google Scholar]
18. Köhler S., Carmody L., Vasilevsky N., Jacobsen J.O.B., Danis D., Gourdine J.-P., Gargano M., Harris N.L., Matentzoglu N., McMurry J.A. et al. .. Expansion of the Human Phenotype Ontology (HPO) knowledge base and resources. Nucleic Acids Res. 2019; 47:D1018–D1027. [Europe PMC free article] [Abstract] [Google Scholar]
19. Karczewski K.J., Francioli L.C., Tiao G., Cummings B.B., Alföldi J., Wang Q., Collins R.L., Laricchia K.M., Ganna A., Birnbaum D.P. et al. .. Variation across 141,456 human exomes and genomes reveals the spectrum of loss-of-function intolerance across human protein-coding genes. 2019; 30 January 2019, preprint: not peer reviewed10.1101/531210. [CrossRef]
20. McLaren W., Gil L., Hunt S.E., Riat H.S., Ritchie G.R.S., Thormann A., Flicek P., Cunningham F.. The Ensembl variant effect predictor. Genome Biol. 2016; 17:122. [Europe PMC free article] [Abstract] [Google Scholar]
21. Buniello A., MacArthur J.A.L., Cerezo M., Harris L.W., Hayhurst J., Malangone C., McMahon A., Morales J., Mountjoy E., Sollis E. et al. .. The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. Nucleic Acids Res. 2019; 47:D1005–D1012. [Europe PMC free article] [Abstract] [Google Scholar]
22. Li M.J., Liu Z., Wang P., Wong M.P., Nelson M.R., Kocher J.-P.A., Yeager M., Sham P.C., Chanock S.J., Xia Z. et al. .. GWASdb v2: an update database for human genetic variants identified by genome-wide association studies. Nucleic Acids Res. 2016; 44:D869–D876. [Europe PMC free article] [Abstract] [Google Scholar]
23. Strande N.T., Riggs E.R., Buchanan A.H., Ceyhan-Birsoy O., DiStefano M., Dwight S.S., Goldstein J., Ghosh R., Seifert B.A., Sneddon T.P. et al. .. Evaluating the clinical validity of gene-disease associations: an evidence-based framework developed by the clinical genome resource. Am. J. Hum. Genet. 2017; 100:895–906. [Europe PMC free article] [Abstract] [Google Scholar]
24. Gutiérrez-Sacristán A., Bravo À., Portero-Tresserra M., Valverde O., Armario A., Blanco-Gandía M.C., Farré A., Fernández-Ibarrondo L., Fonseca F., Giraldo J. et al. .. Text mining and expert curation to develop a database on psychiatric diseases and their genes. Database. 2017; 2017:bax043. [Europe PMC free article] [Abstract] [Google Scholar]
25. Dumontier M., Baker C.J., Baran J., Callahan A., Chepelev L., Cruz-Toledo J., Del Rio N.R., Duck G., Furlong L.I., Keath N. et al. .. The Semanticscience Integrated Ontology (SIO) for biomedical research and knowledge discovery. J. Biomed. Semantics. 2014; 5:14. [Europe PMC free article] [Abstract] [Google Scholar]
26. Tinsley J.M., Fairclough R.J., Storer R., Wilkes F.J., Potter A.C., Squire S.E., Powell D.S., Cozzoli A., Capogrosso R.F., Lambert A. et al. .. Daily treatment with SMTC1100, a novel small molecule utrophin upregulator, dramatically reduces the dystrophic symptoms in the mdx mouse. PLoS One. 2011; 6:e19189. [Europe PMC free article] [Abstract] [Google Scholar]
27. Wojtal D., Kemaladewi D.U., Malam Z., Abdullah S., Wong T.W.Y., Hyatt E., Baghestani Z., Pereira S., Stavropoulos J., Mouly V. et al. .. Spell Checking Nature: Versatility of CRISPR/Cas9 for developing treatments for inherited disorders. Am. J. Hum. Genet. 2016; 98:90–101. [Europe PMC free article] [Abstract] [Google Scholar]
28. Slenter D.N., Kutmon M., Hanspers K., Riutta A., Windsor J., Nunes N., Mélius J., Cirillo E., Coort S.L., Digles D. et al. .. WikiPathways: a multifaceted pathway database bridging metabolomics to other omics research. Nucleic Acids Res. 2018; 46:D661–D667. [Europe PMC free article] [Abstract] [Google Scholar]
29. Xue A., Wu Y., Zhu Z., Zhang F., Kemper K.E., Zheng Z., Yengo L., Lloyd-Jones L.R., Sidorenko J., Wu Y. et al. .. Genome-wide association analyses identify 143 risk variants and putative regulatory mechanisms for type 2 diabetes. Nat. Commun. 2018; 9:2941. [Europe PMC free article] [Abstract] [Google Scholar]
30. Pilgaard K., Jensen C.B., Schou J.H., Lyssenko V., Wegner L., Brøns C., Vilsbøll T., Hansen T., Madsbad S., Holst J.J. et al. .. The T allele of rs7903146 TCF7L2 is associated with impaired insulinotropic action of incretin hormones, reduced 24 h profiles of plasma insulin and glucagon, and increased hepatic glucose production in young healthy men. Diabetologia. 2009; 52:1298–1307. [Abstract] [Google Scholar]
31. Villareal D.T., Robertson H., Bell G.I., Patterson B.W., Tran H., Wice B., Polonsky K.S.. TCF7L2 variant rs7903146 affects the risk of type 2 diabetes by modulating incretin action. Diabetes. 2010; 59:479–485. [Europe PMC free article] [Abstract] [Google Scholar]
32. Gjesing A.P., Kjems L.L., Vestmar M.A., Grarup N., Linneberg A., Deacon C.F., Holst J.J., Pedersen O., Hansen T.. Carriers of the TCF7L2 rs7903146 TT genotype have elevated levels of plasma glucose, serum proinsulin and plasma gastric inhibitory polypeptide (GIP) during a meal test. Diabetologia. 2011; 54:103–110. [Abstract] [Google Scholar]
Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

Full text links 

Read article at publisher's site: https://doi.org/10.1093/nar/gkz1021

Read article for free, from open access legal sources, via Unpaywall: https://academic.oup.com/nar/article-pdf/48/D1/D845/31697865/gkz1021.pdfUnpaywall

Citations & impact 

Impact metrics

1,108
Citations
Jump to Citations

Citations of article over time

Alternative metrics

Altmetric item for https://www.altmetric.com/details/69746871
Altmetric
Discover the attention surrounding your research
https://www.altmetric.com/details/69746871

Smart citations by scite.ai
Smart citations by scite.ai include citation statements extracted from the full text of the citing article. The number of the statements may be higher than the number of citations provided by EuropePMC if one paper cites another multiple times or lower if scite has not yet processed some of the citing articles.
Explore citation contexts and check if this article has been supported or disputed.
https://scite.ai/reports/10.1093/nar/gkz1021

Supporting
Mentioning
Contrasting
6
1571
0

Article citations

Go to all (1,108) article citations

Data 

Data behind the article

This data has been text mined from the article, or deposited into data resources.

BioStudies: supplemental material and supporting data

Diseases

SNPs (3)

Similar Articles 

To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.

Funding 

Funders who supported this work.

Agència de Gestió d’Ajuts Universitaris i de Recerca Generalitat de Catalunya

    EU H2020 Programme (1)

    EU-FP7 (1)

    FEDER (1)

    IMI-JU

      ISCIII

        ISCIII-FEDER (3)

        MINECO (1)

        Research Programme on Biomedical Informatics

          Spanish National Bioinformatics Institute