1-20 of 8952
Sort by
Journal Article
Wilailack Meesawad and others
Database, Volume 2025, 2025, baae127, https://doi.org/10.1093/database/baae127
Published: 22 May 2025
Image
Published: 22 May 2025
Figure 1. Model architecture.
Image
Published: 22 May 2025
Figure 2. Workflow of our approach.
Image
Published: 20 May 2025
Figure 1. Screenshot of the Molgenis catalogue’s class hierarchy in Graph2VR. The instances of the “subcohorts” are expanded.
Image
Published: 20 May 2025
Figure 7. (a) GraphDB layout after manually ordering the nodes. An overlay was added to distinguish which nodes belong to which dataset. The different colours indicate that the nodes have different rdf:types (e.g. light blue being a gene and dark blue a “Disease, Disorder or Finding”). However, not all data w
Image
Published: 20 May 2025
Figure 2. An example figure showing the selection of the topic “tobacco,” followed by navigation to the “DNBC” cohort. The users then specified a query pattern and saw that the PSYCONN cohort was also linked to the same topic in this example set (the original catalogue holds many more records). (The labels we
Image
Published: 20 May 2025
Figure 3. How the user made a query pattern for the variable “blood pressure” and then found all variable-mappings that fulfil this condition. There are three cohorts (G21, ENVIRONAGE, and GENR) in this sample dataset that match, as seen in the middle diamond. This example data was taken from the LifeCycle pr
Image
Published: 20 May 2025
Figure 6. (a) Graph2VR layout. (b): GraphDB default layout. Red nodes represent patients. Blue nodes represent phenotypes. Clustering of individuals with the same phenotype (minimum of two per HPO) in Graph2VR 3D force-directed layout (a) and in GraphDB automatic layout (b).
Image
Published: 20 May 2025
Figure 10. Example of a large graph demonstrating the rendering capabilities of Graph2VR on an RTX 4090. It shows about 17 000 nodes handled by Graph2VR. A framerate drop to 14 fps was observed during the layout phase (the fps increase significantly afterwards). Dataset: DBpedia.
Journal Article
Alexander J Kellmann and others
Database, Volume 2025, 2025, baaf008, https://doi.org/10.1093/database/baaf008
Published: 20 May 2025
Image
Published: 20 May 2025
Figure 4. This figure provides a schematic overview of the structure and interconnections within the RDF data (turtle files) generated from VIP VCF files including potential references to external sources. Each box represents a node. Each line represents a predicate. The arrow points towards the object of a t
Image
Published: 20 May 2025
Figure 5. (a) Graph2VR using automatic layouts (first Class hierarchy layout, then 3D Force directed). Even though the graphs in the screenshot look very small, they are only spread out, and the user can easily navigate closer in VR to interact with them. (b) GraphDB manual layout where the red nodes represen
Image
Published: 20 May 2025
Figure 8. The graph resulting from the query is based on all three graphs. It connects the entries about the ATC codes of aspirin from Drugbank with the ATC code ontology, connects the Drugbank entry for aspirin to PubChem, and displays its CAS number(s).
Image
Published: 20 May 2025
Figure 9. The graph from the previous query about Aspirin. It has been expanded using data from Drugbank and stretched manually. At left are the CAS numbers from PubChem, in the middle the expanded Drugbank entry, and on the right the ATC code from the third database, demonstrating that this can be used as a
Image
Published: 09 May 2025
Figure 4. The glycolysis pathway. Enzymes are represented by light blue boxes; yellow boxes correspond to intermediate molecules within the pathway. Colourful boxes flanking the pathway represent the DA of each enzyme where each SCOP domain is represented by the superfamily code. SF 53067: c.55.1 (actin-like
Image
Published: 09 May 2025
Figure 5. Mutations found in LOG genes from the Solanum genus mapped on CRMP (LOG8) from A. thaliana (PDB: 1YDH). Red represents location of deletions, yellow represents location of insertions, and green marks the flanking residues for intronic mutations. The orange area marks the inferred substrate bindi
Journal Article
Sarthak Joshi and others
Database, Volume 2025, 2025, baaf035, https://doi.org/10.1093/database/baaf035
Published: 09 May 2025
Image
Published: 09 May 2025
Figure 1. GenDiS3 workflow for search and validation of homologs.
Image
Published: 09 May 2025
Figure 2. Frequency distribution of superfamilies with respect to true positive rates of the search strategy. The dashed lines indicate the 5th, 15th, 25th, 50th, and 75th percentiles of the data. From the plot, we can observe that 50% of the superfamilies have a true positive rate of >0.77 for DELTA-BLAST
Image
Published: 09 May 2025
Figure 3. The GenDiS3 web interface for visualizing taxonomic distribution and DAs. (a) Sunburst chart (interactive) showing the taxonomic distribution of a superfamily. (b) SCOP DAs of a superfamily.