Interpretation of edge (arrow) and node colors.

[fenghuijian]

How should the colors of the edges and nodes be interpreted? If red indicates repression and green indicates activation, then why is transcription shown in green — what kind of regu

latory relationship does that represent? #

[pablormier]

Hi @fenghuijian, could you give us more details about which particular method are you running, and code for reproducing if possible? @vicpaton @deeenes

[fenghuijian]

I would appreciate it if you could provide a detailed explanation of the visual elements presented in the first figure from your tutorial. Specifically, could you clarify the biological interpretation of edge colors—do they represent activating or inhibitory interactions? Furthermore, what do the node colors indicate in terms of gene expression levels (e.g., high vs. low expression)? What is the significance of gray-colored nodes, and how should node size be interpreted?

In addition, could you offer a more comprehensive interpretation of the network shown in the first figure? For example, how does EGFR influence the expression of the FOSB gene? Does EGFR function as an activator or an inhibitor in this context?

[fenghuijian]

Codes ：https://networkcommons.readthedocs.io/en/latest/vignettes/2_multiple_methods.html
source_df = pd.DataFrame({'source': ['EGFR'],'sign': [-1]}, columns=['source', 'sign'])
source_df.set_index('source', inplace=True)
sources = source_df['sign'].to_dict()

shortest_sc_network, shortest_sc_list = nc.methods.run_sign_consistency(shortest_path_network, shortest_paths_list, sources,measurements)
visualizer = nc.visual.NetworkXVisualizer(shortest_sc_network)
visualizer.visualize_network(sources, measurements, network_type='sign_consistent')
Upon selecting network_type='sign_consistent', the resulting network graph displays edges and nodes in varying shades of red and green. Could you please explain how these color representations should be interpreted in a biological context?

[fenghuijian]

One more question: When I run this line of code，how can I extract the structured information from the graph?
visualizer = nc.visual.NetworkXVisualizer(shortest_sc_network)
visualizer.visualize_network(sources, measurements, network_type='sign_consistent')

[fenghuijian]

Thank you very much for your detailed response！

[vicpaton]

@fenghuijian thanks for your question. In your first post, the signs make sense because your initial node is down regulated (-1, see your code source_df = pd.DataFrame({'source': ['EGFR'],'sign': [-1]}, columns=['source', 'sign']). In your network, the edge product between EGFR and FOSB is -1, but since EGFR is already downregulated, -1*-1 is 1, that's why FOSB is upregulated in the network. The method is working as intended but it is true that the visualization can be clearer, we'll think about how to improve it.

[fenghuijian]

Thank you for your helpful responses. This is a great tool, but it lacks sufficient explanation or documentation for signal network.

[pablormier]

@fenghuijian networkcommons offers general interfaces for many methods, but if you want to get more into details about intracellular signaling network modeling, please check https://corneto.org which is the tool used by NetworkCommons for some of the methods included there. See for example https://corneto.org/dev/guide/signaling/carnival.html and https://corneto.org/dev/tutorials/single-sample-carnival-transcriptomics.html.

CORNETO focuses on network modeling, whereas NetworkCommons focuses on facilitating access to many different methods (the ones included in CORNETO plus others not related to CORNETO formalism) for application to datasets and benchmarking.