Hello,
I ran MultiNicheNet and in some cases, I find interactions that exist in the 2 directions, with the 2 genes involved being both ligand or receptor. I went to the database to explore these pairs and where they were coming from, and I saw that it is often that in one direction, it come from OmniPath, and the other direction comes from Verschueren et al. To check if it is contradictory information or just an interaction that exists in both directions, I went to the Verschueren paper. Their interactions can be visualized there:
link
I saw in the supplementary note that you got the interactions from the supp. table 2. I guess you picked the interactions from IGSF_interactome_predictions sheet which are characterized as 'positive' with high confidence. The thing is that the Cytoscape network they provide and the table sometimes don't match.
For example: in the table we can find NEGR1-NTM (POS-HICONF) whereas in the Cytoscape network it is NTM-NEGR1. And the Cytoscape direction matches the OmniPath directionality in MNN LR network, and on the contrary the Verschueren reported pair in MNN LR network matches the supp table.
Some other pairs match between the Verschueren table and Cytoscape, like CD2-NLGN3, but the latter is reported as NLGN3-CD2 in the MNN LR network (whereas it is CD2-NLGN3 for OmniPath). Another example of confusion is LRRC4C-LDLR in the MNN database like in the original supp table, but which is LDLR-LRRC4C on the Cytoscape network. Sorry if it is not very clear, it is a bit difficult to explain all the possibilities.
This is quite confusing to me and I struggle knowing which pairs I can really trust. Would you have some insights on that matter ?
Thank you very much,
Best,
Emma
Hello,
I ran MultiNicheNet and in some cases, I find interactions that exist in the 2 directions, with the 2 genes involved being both ligand or receptor. I went to the database to explore these pairs and where they were coming from, and I saw that it is often that in one direction, it come from OmniPath, and the other direction comes from Verschueren et al. To check if it is contradictory information or just an interaction that exists in both directions, I went to the Verschueren paper. Their interactions can be visualized there:
link
I saw in the supplementary note that you got the interactions from the supp. table 2. I guess you picked the interactions from IGSF_interactome_predictions sheet which are characterized as 'positive' with high confidence. The thing is that the Cytoscape network they provide and the table sometimes don't match.
For example: in the table we can find NEGR1-NTM (POS-HICONF) whereas in the Cytoscape network it is NTM-NEGR1. And the Cytoscape direction matches the OmniPath directionality in MNN LR network, and on the contrary the Verschueren reported pair in MNN LR network matches the supp table.
Some other pairs match between the Verschueren table and Cytoscape, like CD2-NLGN3, but the latter is reported as NLGN3-CD2 in the MNN LR network (whereas it is CD2-NLGN3 for OmniPath). Another example of confusion is LRRC4C-LDLR in the MNN database like in the original supp table, but which is LDLR-LRRC4C on the Cytoscape network. Sorry if it is not very clear, it is a bit difficult to explain all the possibilities.
This is quite confusing to me and I struggle knowing which pairs I can really trust. Would you have some insights on that matter ?
Thank you very much,
Best,
Emma