Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Display neighborhood #438

Closed
logmosier opened this issue Feb 15, 2018 · 7 comments
Closed

Display neighborhood #438

logmosier opened this issue Feb 15, 2018 · 7 comments
Assignees

Comments

@logmosier
Copy link
Collaborator

Right now the neighborhood graph that is being fetched, are so large the graph is unreadable. The smallest I have found is 295 nodes and 856 edges.
screen shot 2018-02-15 at 11 32 27 am
But they can get much larger (1572 nodes 15752 edges) to the point of taking 3+ min to display.
screen shot 2018-02-15 at 11 40 06 am
(both of these are limited to 4 interaction types)
Would the best way to handle this be by modifying the data that is fetched (cut away nodes before displaying) or is there a better way?

@logmosier logmosier self-assigned this Feb 15, 2018
@jvwong
Copy link
Member

jvwong commented Feb 15, 2018

I assume this is a client-side bottleneck. Is it possible to spend time finding an example query that returns fewer interactions?

@logmosier
Copy link
Collaborator Author

I was able to find some much smaller graphs, Q8IWf7 and Q9BQY4 are good.

@maxkfranz
Copy link
Member

maxkfranz commented Feb 16, 2018

You can load and render 20k elements in about a second (maybe 2s on a slower machine).

Here are some tests done back on [email protected]: https://github.com/cytoscape/cyjs2.6-cyjs2.7-desktop-comparison And 3.2 is faster than 2.7.

Try using the Chrome profiler to see what's going on. Or use console.time() to track how long different parts of the code are taking.

I suspect you have a slow layout and/or the PC query is taking a while. You can also try using the grid layout or the random layout to see how much of a difference layout expense makes -- those are cheap layouts.

@maxkfranz
Copy link
Member

What layout are you using now?

@logmosier
Copy link
Collaborator Author

I was using the cose-bilkent as I was reusing baseNetworkView, I switched over to grid layout and saw a drop form 26512.44ms to 1279.80ms for the layout.run() call.

@IgorRodchenkov
Copy link
Member

Or try PATHSBETWEEN graph query instead of neighborhood (with one ID is ok)...

@jvwong jvwong closed this as completed Mar 7, 2018
@IgorRodchenkov
Copy link
Member

I bet this (that a PC neighborhood query produces a large "hairball") has to do with
PathwayCommons/cpath2#284 and BioPAX/Paxtools#34 - both will be fixed in the next PC db release (PC10).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants