Database form clarification

[Tianyi409]

Hi Daniel,

Thank you for your package, it is very helpful for studying intercellular communication. I tried to add my dataset to liana for analysis before but failed many times. Luckily, I finally found out why. Because my data comes from mice, and the databases you analyze are all human databases. IN THIS CASE, the case letters for the same gene are distinguished from the mouse gene "Gata3" and the human gene "GATA3". Therefore, I recommend that you clarify this fact in the tutorial. Also, could you please update the mouse database?

PS, I prefer Death Knight to Paladin

[dbdimitrov]

Hi @Tianyi409,

I recommend that you clarify this fact in the tutorial.

Thanks, this is good feedback. I have also noticed that this is a common mistake, I will try to make it more obvious once we update LIANA to LIANA+.

could you please update the mouse database?

In what sense? Though, I agree I should update both the human and mouse consensus :)

PS, I prefer Death Knight to Paladin

Hahaha, all good as long as you choose the correct side (horde) :smile:

[Tianyi409]

My mistake, I have seen your mouse consensus in Resource, which I dont use before. Maybe you can mention it in tutorial, Cuz in the tutorial Consensus is only the Human database

Additionally, can I ask you some meaning of the colname which the LIANA give to me? The "aggregate rank", "Mean Rank" and "LRscore"?

For the Horde!

[dbdimitrov]

aggregate_rank would be a probability distribution calculated with the RobustRankAggregate method of the ranks across different ligand-receptor methods. Mean rank is just the mean rank across different ligand-receptor methods, while LRscore is an expression magnitude score as calculated by SingleCellSignalR :)