RaMP-DB 2.0: a renovated knowledgebase for deriving biological and chemical insight from metabolites, proteins, and genes

Abstract

Motivation: Functional interpretation of high-throughput metabolomic and transcriptomic results is a crucial step in generating insight from experimental data. However, pathway and functional information for genes and metabolites are distributed among many siloed resources, limiting the scope of analyses that rely on a single knowledge source.

Results: RaMP-DB 2.0 is a web interface, relational database, API and R package designed for straightforward and comprehensive functional interpretation of metabolomic and multi-omic data. RaMP-DB 2.0 has been upgraded with an expanded breadth and depth of functional and chemical annotations (ClassyFire, LIPID MAPS, SMILES, InChIs, etc.), with new data types related to metabolites and lipids incorporated. To streamline entity resolution across multiple source databases, we have implemented a new semi-automated process, thereby lessening the burden of harmonization and supporting more frequent updates. The associated RaMP-DB 2.0 R package now supports queries on pathways, common reactions (e.g. metabolite-enzyme relationship), chemical functional ontologies, chemical classes and chemical structures, as well as enrichment analyses on pathways (multi-omic) and chemical classes. Lastly, the RaMP-DB web interface has been completely redesigned using the Angular framework.

Availability and implementation: The code used to build all components of RaMP-DB 2.0 are freely available on GitHub at https://github.com/ncats/ramp-db, https://github.com/ncats/RaMP-Client/ and https://github.com/ncats/RaMP-Backend. The RaMP-DB web application can be accessed at https://rampdb.nih.gov/.

Supplementary information: Supplementary data are available at Bioinformatics online.

Fig 1.

Overview of RaMP-DB. (A) Relationships between metabolites/gene data, annotation types and enrichment options in RaMP. Certain annotations, such as chemical structure, are only available for metabolites. (B) RaMP data sources, and relationships between different components of RaMP, including links to publicly accessible code

Fig 2.

Approach to entity resolution of names across different database sources. The figure depicts the mappings of IDs and metabolites from two different source databases (e.g. denoted as Source 1 and Source 2 Dictionary). Source 2 Dictionary has two instances of ID C, mapping to two different metabolites (1 and 3). In this case, Metabolites 1 and 3 will be merged and considered the same metabolite, which may or may not be accurate

Fig 3.

Overlap in content among source databases. Only analytes mapping to pathways are considered, as HMDB contains a large number of metabolites associated only with ontologies, which are not relevant to Reactome and Wikipathways as pathway-centric databases. (A) Overlap in metabolites associated with at least one pathway between source databases in RaMP. (B) Overlap of genes associated with at least one pathway. The filled circle(s) underneath each bar in the plots demonstrate the source databases that the analyte counts are drawn from

Fig 4.

Screenshots of the updated RaMP-DB web interface. (A) The web interface implements all the new features of the RaMP-DB 2.0 package. A new, more streamlined splash screen organizes the functionalities of the interface into queries for pathways, chemical properties, reactions and ontologies, as well as enrichment analyses. (B) Example pathway enrichment ‘lollipop’ plot generated in the web interface (cropped), using an example query of 20 metabolites and 4 genes. The new plotting function simplifies the process of identifying functional redundancies in pathway analysis output