search

thpar / diffany

Diffany is an open-source toolbox for calculating and visualizing differential networks.
7 stars 2 forks source link

[CLOSED] visualisation of node attributes #200

Closed diffany-admin closed 9 years ago

diffany-admin commented 9 years ago

Issue by svlandeg Fri Oct 17 10:12:35 2014

cf. #198, needed for osmotic use-case: phosphorylation_site (e.g. diamond shaped node) and kinase_function (border around node?)

TODO also in CyNetworkBridge.getNetwork

diffany-admin commented 9 years ago

Comment by svlandeg Thu Oct 23 13:54:50 2014

See also #195 for DE node information that is encoded as edge properties: needs to be incorporated into the cytoscape plugin as well (let's discuss this first)

diffany-admin commented 9 years ago

Comment by tvparys Mon Nov 10 08:48:27 2014

@svlandeg

I can not encode this as node properties, because our plugin requires all networks (the reference and all time-specific networks) to be within the same network collection, and node properties can not vary between networks in Cytoscape.

Apparently: yes they can! This could be version dependent, but my current tests show that each single node in Cytoscape 3.1.1 (even within the same network collection) has a unique SUID, so node attributes can be set for each node separately. If the plugin is changed to support this. Even copy/pasting or cloning networks or nodes, makes exact copies of the nodes, except for its SUID.

Currently our networks now have the attributes:

diffany-admin commented 9 years ago

Comment by tvparys Mon Nov 10 12:12:50 2014

@svlandeg : can the attributes phosphorylation_site and kinase_function change according to the network?

diffany-admin commented 9 years ago

Comment by tvparys Mon Nov 10 12:37:06 2014

More importantly, how tight are those extra attributes tied to the Diffany algorithm? In other words: does it make sense for the plugin to explicitly mention these attributes? If not, it suffices to make Diffany Plugin retain node attributes over Network to CyNetwork transitions and vice versa.

Visualization of use case specific properties should then be up to the user to add them to the VizMapper.

diffany-admin commented 9 years ago

Comment by tvparys Mon Nov 10 15:19:45 2014

What happens now if nodes with the same name have different values in the different input networks?

diffany-admin commented 9 years ago

Comment by svlandeg Wed Nov 12 07:23:40 2014

the phosphorylation site and the kinase function attributes are invariable per protein/gene, or in other words for each unique Diffany node it will be invariable. Currently this information is not used in the algorithm, but it could be at some point, as these attributes might help zooming in to a relevant part of the network, for instance

diffany-admin commented 9 years ago

Comment by svlandeg Wed Nov 12 07:25:50 2014

I obviously like the generic approach of keeping all node attributes in the network transitions, but I would still add the phosphorylation/kinase data to the default vizmapper. Because this will be included in the example networks we'll package with Diffany, and helps showcase the functionality that is possible with the tool.

diffany-admin commented 9 years ago

Comment by tvparys Wed Nov 12 14:52:02 2014

included in 32c1c22b71a19b0b0beb52ee7bf6bfb6d1430db0