how to set parameters in TO-GCN

[cwu2011]

In the pipeline, TO-GCN is called by following command
TO-GCN 13 13 example_data/TFs_1718.tsv example_data/TFs_1718.tsv 0.84 0.84 example_data/seeds.txt 0
when we want to run the TO-GCN program to assign the time-order (level) of nodes in C1+C2+ GCN.

if we want to assign the time-order of node s in C1+C2- , C1-C2+ or C10 C2-, how to set the parameters of cutoff and co-expression type?
should they be set as 0.84 -0.75(-0.75, 0.84), and a co-expression type number that not mentioned in the pipeline document?

thanks!

[petitmingchang]

@cwu2011 For generating the time ordered with "TO-GCN" command, we only provide the coexpresion types of C1+C2+, C1+C20, and C10C2+ so far. Because we found it is difficult to interpret the meaning of networks with other coexpression types.

[cwu2011]

thanks for your answering.
There are 2 other questions:

1. In the paper you selected a gene with a monotonic decreasing trend both in C1 and C2 in the case of C1+C2+ .
   Can i choose seed genes in the same way for the cases of C1+C20 and C10C2+, or choose seed genes that only monotonic decrease in C1 or C2?
2. Whats the difference between TO-GCN.exe and TO-GCN2.exe? which one is recommended to use? What kind of scenarios do they apply to?

[petitmingchang]

@cwu2011

1. If you would like to use C1+C20, which means we do not care the gene coexpression relationship in C2, then just choose the monotonic descending gene in C1. Similarly, if you would like to use C10C2+ then just consider the monotonic descending gene in C2.
2. Yes, TO-GCN2 is an extended version of TO-GCN that can generate networks with C1+C2- and C1-C2+ coexpression types by setting the last parameter to 3 and 4, respectively. But as I mentioned previously, you have to interpret the meanings of these networks cautiously.

[cwu2011]

thank you very much!