Network containing fixed/locked nodes doesn't arrange as expected

[olga.s...@gmail.com]

Hi there,

First of all, thanks for creating and maintaining this fantastic software! (I am working with the latest desktop Cytoscape version (3.7.2))

I would like to create a network similar to one that I saw in a paper (see attached, Brockmann et al., Nature 2017, Figure 2B).

My network involves 11 fixed nodes (red) arranged in a circle surrounding a few hundred nodes. I managed to lock these outer nodes in circle shape by using the bypass feature for X and Y positions in Styles. Now I would like for the nodes in the middle - each of which connects to at least 2 of the fixed outer nodes - to be arranged so that I can easily see groups of nodes that connect to the same two or three outer nodes. E.g., the nodes that connect to two neighboring outer nodes should all be clustered together between those two fixed nodes. I am surprised that none of the layouts gets me there since I would have thought that this is what would minimize the total edge-length in the network (see an example of my unsuccessful attempts attached - fixed nodes in red). Perhaps the fixed nodes prevent the layout algorithms from working as expected? I would be super happy about any insights on why this is not working for me and if there is a way to achieve something close to that published figure besides placing the nodes manually.

(In the published figure, there is a ring of nodes that are connected to only a single fixed (red) node - for me, these are of lower priority for now and I removed these types of nodes from my network.)

Thank you very much in advance!

Olga

[Lars Juhl Jensen]

This sounds like a problem that the boudaryLayout app is designed for. You could constrain the outer nodes each within their own little circle and all other nodes within a circle in the middle.

BW

Lars

[Alex Pico]

BoundaryLayout app

 * App page: http://apps.cytoscape.org/apps/boundarylayout

 * Documentation: http://www.rbvi.ucsf.edu/cytoscape/boundaryLayout/index.shtml

 - Alex

--
You received this message because you are subscribed to the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cytoscape-helpd...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/5965e032-c78d-4471-867d-1beba605486f%40googlegroups.com.

[Olga Schubert]

Dear Lars and Alex,

Thanks so much for your advice to use the BoundaryLayout tool. I played around with it today and it seems to be doing mostly what I wanted (see attached) but there is some behavior that I don't understand yet.

For example (for simplicity I reduced the shared nodes in the inner circle to only ones that connect to two outer nodes), I would expect that between each of two neighboring nodes there is a small accumulation of shared nodes (if there are any), however, sometimes these shared nodes are much further away in the inner circle. Shouldn't they always be located automatically in a way that the sum of their edges are minimal? Could it be that this is just because I didn't set the parameters properly? Or how would I achieve this?

Thank you again!

Olga

You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/9Jn7l7wXc-k/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/EC2D0431-7C33-4BD5-89AA-026BF7AFF79B%40gladstone.ucsf.edu.

[Scooter Morris]

Hi Olga,

   I would definitely try to tune the parameters.  BoundaryLayout is attempting to solve a relatively complex problem, so the way it optimizes the layout might not be ideal for you.  Also, keep in mind that if the nodes in the middle are connected to more than one external node they will naturally move towards the middle.  You can tune that behavior by assigning edge weights and then using those weights in the layout, but without seeing your network, I can't tell for sure if that will work.

-- scooter

On Tuesday, March 24, 2020 at 10:19:06 PM UTC-7, Olga Schubert wrote:

Dear Lars and Alex,

Thanks so much for your advice to use the BoundaryLayout tool. I played around with it today and it seems to be doing mostly what I wanted (see attached) but there is some behavior that I don't understand yet.

For example (for simplicity I reduced the shared nodes in the inner circle to only ones that connect to two outer nodes), I would expect that between each of two neighboring nodes there is a small accumulation of shared nodes (if there are any), however, sometimes these shared nodes are much further away in the inner circle. Shouldn't they always be located automatically in a way that the sum of their edges are minimal? Could it be that this is just because I didn't set the parameters properly? Or how would I achieve this?

Thank you again!

Olga

On Tue, Mar 24, 2020 at 11:22 AM Alex Pico <alex...@gladstone.ucsf.edu> wrote:

BoundaryLayout app

 * App page: http://apps.cytoscape.org/apps/boundarylayout

 * Documentation: http://www.rbvi.ucsf.edu/cytoscape/boundaryLayout/index.shtml

 - Alex

On Mar 23, 2020, at 11:32 PM, Lars Juhl Jensen <lars.ju...@gmail.com> wrote:

This sounds like a problem that the boudaryLayout app is designed for. You could constrain the outer nodes each within their own little circle and all other nodes within a circle in the middle.

BW

Lars

--
You received this message because you are subscribed to the Google Groups "cytoscape-helpdesk" group.

To unsubscribe from this group and stop receiving emails from it, send an email to cytoscape-helpdesk+unsub...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/5965e032-c78d-4471-867d-1beba605486f%40googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/9Jn7l7wXc-k/unsubscribe.

To unsubscribe from this group and all its topics, send an email to cytoscape-helpdesk+unsub...@googlegroups.com.