What it does
A gene regulatory network lists which genes (usually transcription factors) switch which other genes on or off. As a list it is hard to read; as a picture with a few numbers, its structure jumps out: the hubs that regulate many genes, the genes that switch themselves off, and the small recurring wiring patterns called network motifs.
This page reads a network in SIF or GraphML, draws it, and works out:
- In-degree and out-degree of every gene: how many regulators it has and how many genes it regulates, split into activation and repression.
- Density, the share of all possible links that exist, and the connected parts of the network.
- Autoregulation: genes that regulate themselves, positive or negative.
- Two-gene feedback loops: pairs that regulate each other.
- Feed-forward loops: X regulates Y, and both regulate Z. They are coherent when the direct path has the same sign as the path through Y, and incoherent when the signs differ. The best-known kinds are named: C1 (all activation, a sign-sensitive delay) and I1 (X activates Y and Z, Y represses Z, a pulse generator).
Nothing is uploaded; the file is read and drawn in your browser.
How to use it
- Paste a network, or press Open a file to read a .sif or .graphml file exported from Cytoscape, yEd, NetworkX or igraph. In SIF each line is
source interaction target, with more targets allowed on one line; a name alone is a gene with no links. - Choose the Layout: force-directed (linked genes close together) or a circle. New layout tries another start if genes overlap.
- Highlight feed-forward loops or feedback, so the rest of the network fades.
- Size nodes by number of links or keep them all the same, and switch gene names on or off.
- Read the facts and tables, download the degrees and the motifs as CSV, or the report as an image, PDF or Word file.
Worked example
The example is a made-up network of 13 genes and 17 interactions: 13 activating and 4 repressing. MasterA, switched on by Sensor, regulates TF1, TF2, GeneX, GeneY and GeneZ; TF3 represses itself, TF4 activates itself, and TF4 and TF5 form a negative two-gene loop. Orphan has no links.
The page finds a density of 0.109, three parts (9, 3 and 1 genes) and 4 feed-forward loops: MasterA → TF1 → GeneX and MasterA → TF1 → GeneY are coherent C1, MasterA → TF2 ⊣ GeneY is incoherent I1, and MasterA → TF2 → GeneZ with MasterA repressing GeneZ is incoherent too. MasterA is the biggest hub, with 6 links (5 out, 1 in). NetworkX gives the same degrees, parts and loops.
Tips
- Interaction words are matched loosely: activates, activation, induces, stimulates, + and 1 are activation; represses, inhibits, suppresses, −, -1 and -| are repression. Others (binds, pp) are drawn as plain lines and count as unsigned.
- A repeated line for the same pair of genes is counted once; the facts say how many were left out.
- Big networks are drawn on a circle above 800 genes; the numbers are worked out for up to 3,000.
FAQ
Are loops counted once? Yes. Each feed-forward loop is one X, Y, Z triple; extra links between the three (such as Z regulating X) do not hide it.
Why does the picture change with New layout? Force-directed layouts start from random positions. The same seed always gives the same picture, so a download matches the screen.
Does my file leave the device? No. It is read with the browser’s own file reader, and the page works offline. For study and planning, not a substitute for professional judgement.
Where the numbers come from
- Simple Interaction Format (SIF) and GraphML as used by Cytoscape and other tools; force-directed drawing after Fruchterman, T. M. J. and Reingold, E. M. (1991), Software: Practice and Experience 21, 1129 to 1164; network motifs and the coherent/incoherent feed-forward loop classes after Mangan, S. and Alon, U. (2003), PNAS 100, 11980 to 11985: used as methods, written by fibub.
- Checked against NetworkX 3 (degrees, density, weakly connected components, GraphML written by write_graphml, and feed-forward loops counted with DiGraphMatcher) on the example and on a random 40-gene network. The example network is made up.
Standard scientific values, such as constants, atomic weights and melting points, are facts found in every textbook; we name where ours were checked. If you are an author or publisher and want a value, a credit or a table changed or removed, tell us and we will act quickly.
Last updated 2026-11-09.