Species gene-name convention discrepancy

[VittoriaOssanna]

Hi all,

I’m working with a mouse scRNA-seq dataset using CellChat, and I noticed something confusing about the CellChatDB.mouse interaction names.

While exploring the CellChat object, I saw that the interaction names stored in the @net outputs are all uppercase. I wanted to confirm that I was indeed using the mouse database (CellChatDB.mouse), which I am.

For example:

head(CellChatDB.mouse$interaction$ligand.symbol, n = 10)
head(CellChatDB.mouse$interaction$receptor.symbol)

returns:

'Tgfb1' 'Tgfb2' 'Tgfb3' 'Tgfb1' 'Tgfb1' 'Tgfb2' 'Tgfb2' 'Tgfb3' 'Tgfb3' 'Tgfb1'

'Tgfbr1, Tgfbr2'
'Tgfbr1, Tgfbr2'
'Tgfbr1, Tgfbr2'
'Acvr1b, Tgfbr2'
'Acvr1c, Tgfbr2'
'Acvr1b, Tgfbr2'

However:

head(CellChatDB.mouse$interaction$interaction_name)

returns:

'TGFB1_TGFBR1_TGFBR2'
'TGFB2_TGFBR1_TGFBR2'
'TGFB3_TGFBR1_TGFBR2'
'TGFB1_ACVR1B_TGFBR2'
'TGFB1_ACVR1C_TGFBR2'
'TGFB2_ACVR1B_TGFBR2'

My understanding is that interaction_name is essentially a concatenation of the ligand and receptor names, but I’m wondering why these are converted to uppercase, whereas the ligand/receptor symbols themselves retain standard mouse gene capitalization.

Is this uppercase formatting intentional for internal consistency or downstream compatibility within the CellChat pipeline? Or could this indicate that I’m missing something in how the database is structured or used?

Thanks a lot!