Problem with canonicalization of favipiravir

[schatzsc]

When I try to compare the two "hydrogen-pruned" versions of favipiravir, the canonicalization mixes up the two ring nitrogen atoms, which also hampers with the new tautomer-detection routine I am now working on:

favipiravir_1.mol
ChemDraw08042201242D

0 0 0 0 0 0 V3000
M V30 BEGIN CTAB
M V30 COUNTS 15 15 0 0 0
M V30 BEGIN ATOM
M V30 1 C -1.071706 -0.206250 0.000000 0
M V30 2 C -1.071706 -1.031250 0.000000 0
M V30 3 N -0.357235 -1.443750 0.000000 0
M V30 4 C 0.357235 -1.031250 0.000000 0
M V30 5 C 0.357235 -0.206250 0.000000 0
M V30 6 N -0.357235 0.206250 0.000000 0
M V30 7 F -1.786177 0.206250 0.000000 0
M V30 8 H -1.786177 -1.443750 0.000000 0
M V30 9 O 1.071706 -1.443750 0.000000 0
M V30 10 C 1.071706 0.206250 0.000000 0
M V30 11 H 1.786177 -1.031250 0.000000 0
M V30 12 O 1.786177 -0.206250 0.000000 0
M V30 13 N 1.071706 1.031250 0.000000 0
M V30 14 H 1.786177 1.443750 0.000000 0
M V30 15 H 0.357235 1.443750 0.000000 0
M V30 END ATOM
M V30 BEGIN BOND
M V30 1 2 1 2
M V30 2 1 2 3
M V30 3 2 3 4
M V30 4 1 4 5
M V30 5 2 5 6
M V30 6 1 6 1
M V30 7 1 1 7
M V30 8 1 2 8
M V30 9 1 4 9
M V30 10 1 5 10
M V30 11 1 9 11
M V30 12 2 10 12
M V30 13 1 10 13
M V30 14 1 13 14
M V30 15 1 13 15
M V30 END BOND
M V30 END CTAB
M END

[JanCBrammer]

@schatzsc, can you elaborate on what you did to obtain the middle and bottom figures? What was the sequence of steps resulting in those figures?

[schatzsc]

Test was run with tautomer.py from

https://github.com/schatzsc/TUCAN_playground/blob/main/tautomer.py

Functions run on both graphs after output of top figure from that file were:

set_hydrogens_count(m1_sorted)
prune_hydrogens(m1_sorted)
nx.is_isomorphic(m1_sorted, m2_sorted)

lines 68, 91, and 111 of tautomer.py for drawing
favipiravir_1.mol and favipiravir_2.mol input files now uploaded to \test\molfiles together with two more tests for tautomerism, amino_acid_1/2.mol and tautomer1_amide.mol vs. tautomer2_enol.mol

[JanCBrammer]

It seems that, in order to reproduce the bug, we'd need access to code that's currently not in this repository. Closing this for the moment since I cannot access https://github.com/schatzsc/TUCAN_playground, it seems to be private or gone.

[schatzsc]

I sent you an invitation to my "TUCAN_playground" but it possibly expired during holidays, so just send it again - can you access it now?