Targetix #71
Targetix #71
Conversation
…, that the rankings of the edges and genes will be restricted to the subnetwork of genes corresponding to regix and targetix, and should be interpreted in that context.
Rebasing targetix to match current master
There was a problem hiding this comment.
We should update the main readme to describe the new optional targetix. Even if we don't add a full test case, it would also be good idea to make a copy of data1 with the targetix set showing how to use it, for example adding a new X_targetix_test.mat.
I'll work on updating the Docker image to see if our test cases still pass with this version of the code.
code/SINGE_GLG_Test.m
Outdated
| if iscolumn(targetix) | ||
| targetix = targetix'; | ||
| elseif ~isrow(targetix) | ||
| display('Error: Regulator indices must be a vector'); |
There was a problem hiding this comment.
Do we need to stop execution here instead of only displaying a message? Or notify the user we are using all genes because their target index was improperly formatted? Currently we won't hit the code in the else block that sets the target index to be all genes.
Also, I believe we want to say "Target indices" not "Regulator indices".
There was a problem hiding this comment.
Changed it to Target indices. I think it's best to error out here especially if this were to use expensive compute resources.
This adds the ability to specify target indices using a variable array
targetixinstead of running GLG tests for all genes as targets. This will allow users to query top regulators of a specific gene, or query the interactions in a small set of regulator and target genes in combination with theregixvariable array.