Node attributes output

[iamjli]

-Gene symbol -Prize value -Robustness -Specificity -Protein type (steiner, TF, terminal)

[AmandaKedaigle]

Can I add a couple? -ProteinChange -GeneChange

(This is for the visualization we're working on with Laura - we need a ProteinChange as opposed to prize value so that we can distinguish between up- and down-regulated proteins. i.e. if prizes are abs(logFC) this would be signed logFC. And geneChange would be for showing if that protein was also up- or down- in RNA-seq)

[alexlenail]

@iamjli @AmandaKedaigle Happy to add these. The abstraction I think makes most sense is that if you pass all of these in, I can pass them back out to you in the solution. The prize file has been

name    prize

Now it will be

name    prize    other fields...

And those other fields will always come out on your solution. Does that sound alright?

[alexlenail]

@iamjli please define Robustness, Specificity

[AmandaKedaigle]

• ProteinChange and GeneChange can be input like that (give a node 0 if there is no value), but the others should be calculated.
• NodeType should be Proteomic if input in the prize file, TF if input in the Garnet File, and Proteomic_TF if both, and Steiner if it was not an input node.
• Robustness is how many times the node appears out of the noisyEdge runs.
• Specificity is how many times the node appears out of the randomTeminal runs.

[alexlenail]

This is done, see https://github.com/fraenkel-lab/OmicsIntegrator2/commit/08829859da041afc13f1d1f001af3413f9b6ad2f