If an input to a binding reaction contains the enabler, remove the corresponding input from the MF

[vanaukenk]

From the 2023-10-12 Reactome2GO meeting at NYU:

In the nuclear signaling by ERBB4 model, there is an example where STAT5 enables a DNA-binding transcription factor activity, but STAT5 is also a member of a protein-containing complex that is an input to that same activity.

In these cases, i.e. when the enabler is part of an input, do not include the protein-containing complex as an input.

[dustine32]

For the STAT5A example:

This one shows how this is a bit more complex than just checking a flat list of complex components. One of the inputs is the complex to remove "ERBB4s80:p-Y694-STAT5A:CSN2 gene" R-HSA-8954223 because the reaction's enabler "STAT5A_HUMAN" R-HSA-507995 is contained many levels down within this complex:

• ERBB4s80:p-Y694-STAT5A:CSN2 gene
  • ERBB4s80:p-Y694-STAT5A
    • STAT5A homodimer
      • p-Y694-STAT5A (R-HSA-507995)

Note: I'm not really sure what's with the difference in labels "STAT5A_HUMAN" vs "p-Y694-STAT5A" but they are both for R-HSA-507995.

Recursive traversal of the complex component hierarchy to check this is pretty trivial but I just wanted to highlight that this will be needed.