Knowledge Network Miner (KnetMiner)[1][2] is an ecosystem of tools to integrate, search and visualise biological Knowledge Graphs (KGs). KnetMiner KGs are built using the open-source data integration platform KnetBuilder and provided in OXL, Neo4j and RDF graph formats. The KnetMiner API provides web endpoints to search the KG with genes and keywords and return the knowledge in graph format. Candidate genes are ranked according to their strength of association with particular search terms. KnetMiner was created and is maintained by researchers at Rothamsted Research.
KnetMiner enables scientists to search across large biological databases and literature to find links between genes, traits, diseases, and many other information types. It makes use of FAIR data[3] principles and supports the use of a range of biological data formats.
KnetMiner is an active project, maintaining a developer and user documentation in its wiki and user manual, respectively. KnetMiner has also been showcased in various press releases[4][5][6][7][8][9]
Social media presence can also be found for KnetMiner on Twitter.
. . . KnetMiner . . .
KnetMiner hosts a range of different species, including a knowledge graph dedicated to SARS-CoV-2[4][9] in response to the 2020 global pandemic, on Rothamsted Research HPC machines, which include the following:
- SARS-CoV-2
- Triticum aestivum
- Oryza Saiva Japonica
- Arabidopsis Thaliana
- Zymoseptoria Tritici
- Fusarium graminearum
KnetMiner has been involved in a number of studies, including studies for wheat,[10][11][12] willow,[12] and SARS-CoV-2.[4] It is also being used for exploring pathogen-host interactions in collaboration with PHI-base, soybean looper, and other species.
KnetMiner contains several core components, including the search interface, the Gene centric view, Evidence centric view, Genomaps.js (chromosome Map view), and KnetMaps.js interactive network view.
The main search view enables users to enter in keyword searches related to any evidence-terms they wish to search for, e.g. “seed dormancy” in Wheat. Complex statements can also be made by including AND, OR, NOT statements.
Gene ID’s can also be provided with or without the search keywords, in the gene list box. This will narrow the search to only these specified gene ID’s. QTL regions can also be specified in the QTL search region box.
The returned genes in Gene View are ranked according to their biological relevance to the search term and/or the volume or evidence present. A user can visualise the genes on a chromosome centric view map in Map View which presents an interactive visualization of high density SNP, QTL, GWAS, and gene data. Map View uses Genomaps.js. The demo for genomaps.js can be found here.
Network View
The Network View [13] uses KnetMaps.js, which is an interactive network visualisation tool for exploration of heterogeneous biological knowledge graphs. A demo of this can be found here and the NPM and GitHub repository here and here, respectively.
. . . KnetMiner . . .