This web-based interface introduces an interactive knowledge graph that is derived from the manuscript: “Machine learning-guided deconvolution of plasma protein levels”, with a series of interactive functions. The interactive knowledge graph provides a dynamic and visual interface to discover biological relationships among five entities: disease, drug, protein, gene, and SNP.
Core features
Knowledge graph visualization
Users can interact with graph using the following options.
Zoom and pan to explore specific regions.
Click on node to inspect its neighbour nodes and the shortest cycle path stated from this node if exsits.
Reset the graph to the initial status by clicking “Reset Selection” button at any time.
Graph filtering
Users can filter the graph based on their interests by selecting specific nodes and edges. The filtering functionality is accessed through three dropdown menus located at the top of the interface.
Important: The dropdown menus should be used sequentially from left to right, as each menu’s available options are dependent on the selection made in the previous one.
Select a network item: Select a node or an edge from the graph.
Select a property: Select the property for the chosen network item
The optional properties for the node:
group: the category of nodes among diseases, drugs, protein, gene, SNP.
id: node id.
label: node label.
The optional properties for the edge:
from (or from_label): constrain the edges from the selected node.
to (or to_label): constrain the edges to the selected node.
Select value(s): Select one or multiple values based on the selected property.
Dynamic updates
Settings for graph filtering immediately update the graph visualization without reloading the page, ensuring a smooth user experience during exploration.
Use case
Visaulization of the neighbour and the shortest cycle path for plasma levels of desmoglein 2 (P::DSG2)
Support and contact
If you encounter any issues or have questions, please contact us via the corresponding author of the publication.
