Breakdown of stereoisomercount in cxcalc wrong, ok in Marvin

User 677b9c22ff

11-10-2008 21:43:16

Hi,

I used all 452458 isomers from C10H16O (via SMOG2).

Code:

cxcalc tetrahedralstereoisomercount doublebondstereoisomercount stereoisomercount C10H16O.sdf > C10H16O-stereoisomercount.txt

or

cxcalc tetrahedralstereoisomercount -m 10000 doublebondstereoisomercount -m 10000 stereoisomercount -m 10000 C10H16O.sdf > C10H16O-stereoisomercount.txt

where m denotes the max number of stereosiomers to be created

however for certain isomers the breakdown is shown like that,

Code:

.. without the -m switch (max stereoisomer)

NUM tetra DBL Totalcount

-----------------------------

10675 0 0 4

10676 0 0 4

10677 0 0 8

10678 0 0 4

10679 0 0 2

10680 0 0 4

10681 0 0 8

10682 0 0 4

10683 0 0 4

10684 0 0 4

10685 0 0 2

10686 0 0 2

10687 0 0 2

10688 0 0 4

10689 0 0 4

10690 0 0 2

10691 0 0 2

10692 0 0 2

10693 0 0 4

10694 0 0 8

10695 0 0 2

10696 0 0 4

10697 0 0 8

10698 0 0 2

10699 0 0 4

10700 0 0 4

10701 0 0 4

10702 0 0 4

10703 0 0 2

10704 0 0 4

10705 0 0 2

10706 0 0 2

10707 0 0 2

10708 0 0 4

10709 0 0 4

10710 0 0 2

10711 0 0 4

10712 0 0 2

10713 0 0 2

10714 0 0 2

10715 0 0 4

10716 0 0 2

10717 0 0 2

10718 0 0 2

10719 0 0 2

10720 0 0 4

Code:

... now with -m switch

NUM tetra DBL Totalcount

-----------------------------

10675 4 1 4

10676 4 1 4

10677 8 1 8

10678 4 1 4

10679 2 1 2

10680 4 1 4

10681 8 1 8

10682 4 1 4

10683 4 1 4

10684 2 2 4

10685 2 1 2

10686 2 1 2

10687 2 1 2

10688 2 2 4

10689 2 2 4

10690 2 1 2

10691 2 1 2

10692 2 1 2

10693 4 1 4

10694 8 1 8

10695 2 1 2

10696 4 1 4

10697 8 1 8

10698 2 1 2

10699 4 1 4

10700 4 1 4

10701 4 1 4

10702 4 1 4

10703 2 1 2

10704 2 2 4

10705 2 1 2

10706 2 1 2

10707 2 1 2

10708 4 1 4

10709 4 1 4

10710 2 1 2

10711 4 1 4

10712 2 1 2

10713 2 1 2

10714 2 1 2

10715 2 2 4

10716 2 1 2

10717 2 1 2

10718 2 1 2

10719 2 1 2

which cant be true because either one needs to be <> 0 if the total count > 0. I don't know what went wrong. A quick

test with bibenzyl "C1=CC=C(C=C1)C1=CC=CC=C1" showed the correct output (1 1 1).

So the max switch should be set higher in general, given the fact that

this calculation is quite performant.

Another one without the explicit max switch

Code:

4972 0 1 4 <---------- ?

4973 4 1 4

4974 3 1 8 <---------- ?

4975 0 1 4

4976 0 0 4 <---------- ?

And one with -m 10000 switch which also reports wrong results

:

Code:

257830 0 1 2

257831 0 1 2

257832 0 1 2

257833 0 1 4 <----- ?

257834 0 1 4 <----- ?

257835 0 1 2

257836 0 1 4

257837 0 1 2

257838 0 1 2

257839 0 0 2

257840 0 0 4

Q1: is there an option "Generate all" for cxcalc?

Q2: How would I invoke the "filter wrong 3D structures" in cxcalc?

Bye

Tobias

ChemAxon e08c317633

21-10-2008 17:01:53

Hi Tobias,

Thanks for the bug report, we will check this error.

Quote:

Q1: is there an option "Generate all" for cxcalc?

Yes, set the "maxstereoisomers" to "-1".

Code:

cxcalc stereoisomers -m -1 mols.sdf

Quote:

Q2: How would I invoke the "filter wrong 3D structures" in cxcalc?

With the " -v, --verify3d" option.

Code:

cxcalc stereoisomers -v mols.sdf

Sorry for the late answer.

Zsolt

User 677b9c22ff

21-10-2008 17:57:30

Thanks,

if the error is not in the API I can also write myself a quick

turnaround.

Tobias

ChemAxon e08c317633

27-10-2008 08:32:55

Hi,

The error is in the API, so we have to fix it.

Zsolt