Benchmarking group assignment against ddbst utility - distinguishing "CH2" from "CY-CH2"

[viniviena]

Hello all. I am benchmarking the program for complicated molecules and comparing with ddbst group assingment utility. The assignment seems to be on point. There is only one assignment that seems to be different as below:

1. GCI Identifier:
   get_groups_from_smiles("CC(=O)OCC1CCC(C23C14C2CC(C3)C4(C)C)(C)C", UNIFACGroups) produces:

("CC(=O)OCC1CCC(C23C14C2CC(C3)C4(C)C)(C)C", ["CH3" => 4, "CH3COO" => 1, "CY-CH2" => 4, "CY-C" => 4, "CY-CH" => 3, "CH2" => 1])

2. using ddbst utility I've got:
   ["CH3" => 4, "CH2" => 5, "CH" => 3, "C" => 4, "CH3COO" => 1 ])

*It seems it is distinguishing "CY-CH2" from "CH2". What is the difference between these two groups? Does it affect the activity coefficient calculation?

[longemen3000]

at first glance CY seems to be a shortcut for Cycle? if that is the case, then it seems to be correct ?

[viniviena]

Oh. Thanks for the clarification. That's seems to be the case. The current algorithm is distinguishing cyclic CH2 from CH2. I don't know if they are separate groups in UNIFAC, but it's nice having an algorithm that distinguish both.