Tabnetviz generates network visualizations from node and edge properties provided in tables. The node and edge properties can be mapped to visual attributes in several ways. Tabnetviz was inspired by the popular Cytoscape program which can also generate similar mappings. However, Cytoscape is a resource-intensive, interactive Java program with a complex graphical interface, and loading networks from tables and defining mappings can be cumbersome in it. Tabnetviz, on the other hand, is a non-interactive, lightweight command-line tool guided by a single text-based configuration file, thus ideal for use in scripts, Makefiles, and reproducible workflows. Once a configuration file has been developed, it only takes a single command to generate the network visualization (typically an SVG file), and to quickly regenerate it whenever the input data changes.
Tabnetviz generates static network visualizations, and is applicable in a wide range of fields such as bioinformatics (for gene regulatory networks, protein interaction networks, etc.), neuroscience, and studies of social networks, computer networks, economic networks, etc.
An example visualization generated by Tabnetviz:
The Tabnetviz configuration file is a YAML format text file, which is easily written manually, and easy to understand. It specifies the node table and the edge table for the network (both can be CSV, TSV, or Excel files), and defines how to map the node and edge properties (provided in node and edge table columns) to visual attributes such as colors, node sizes, shapes, line widths, etc. Node groups and edge groups can also be defined (using Boolean expressions on the node/edge properties), and the mappings can be applied to them.
Tabnetviz is a Python program, and uses Graphviz as its network visualization back-end, and can use any node, edge, and graph attribute known to Graphviz. It also uses Graphviz for generating network layouts. Relying on the power of Graphviz, Tabnetviz can generate high-quality images suitable as illustrations for science publications.
As a bonus, Tabnetviz can optionally calculate numerous graph theoretical quantities such as degrees, centralities, clustering coefficients, etc. These are added to the node/edge table, and can then be mapped to visual attributes, e.g. node sizes or colors.
See the Tabnetviz home page for documentation.