Tom Thomas
Tom Thomas
@earmingol Thank you for the prompt response. The advice to add parameter `device='cuda:0' `worked, when I was using a GPU. However, if i am not using a GPU i.e. fully...
Point re: slower on CPU noted! I had a GPU related question, I am running cell2cell on a GPU, and I am facing memory issues given the size of the...
@earmingol just circled back to tensorcell2cell following trialling a few of the other LR analysis tools. I had to downsample the size to 10% of the dataset due to memory...
facing same issue as above!
@sqjin net.up and net.down appears to look good, at least when I sort by FC, the ligand and receptors do make sense. I think the problem might lie with the...
@sqjin I suppose the other factor here is that there could be a typo in the actual vignette i.e. "#extract the ligand-receptor pairs with upregulated ligands and upregulated recetptors in...
@GBeattie just to clarify the net.up and net.down shouldnt be having overlapping interactions (I think this part is running ok - if you sort by ligand FC etc., it should...
@GBeattie the results i get are more in keeping with the image posted in the issue - where there is significant overlap. [my gene.up list has 397 genes. and my...