Query as text window does not show correct SMARTS?

User 677b9c22ff

02-10-2008 18:55:31

{edit please also see my post from here


SMARTS have cancer}





Hi,


if I do a substructure search in Instant JCHEM and edit SMARTS in MarvinSketch it will also show the substructure as SMARTS in the <Query as text> window.


Something like (Structure Substructure "O1O**O1" ChemTermFilter="")


I can use that SMARTS and also paste it later somwhere else.





In certain cases (currently I have only one) it does not code the correct SMARTS structure and the code itself is different (shorter?).





In case of cyclopropylene structures it will show


(Structure Substructure "*1*=*1" ChemTermFilter="")





In case of Any Atom including H (AH) it will show the same SMARTS


in the text query window, but the correct one in the Sketch Window.


So it would be a cyclopropene with AH in each corner or as SMARTS


[*,#1]1[*,#1]=[*,#1]1





In case of copy/paste it will change the above SMARTS to


[*,#1]-,:-,:1-,:[*,#1]=[*,#1]-,:-,:1





--


If I open the editor it will show the wrong value (because I have the AH


structure in the window. *1*=*1 |$AH_p;AH_p;AH_p$,c:1|


and when I delete it and re-edit it, it will show


[*,#1]1[*,#1]=[*,#1]1 |$AH_p;AH_p;AH_p$,c:1|





So something is wrong here. Maybe two independent errors.


I just wanted to store all my queries from the Text window, but now


I see that I have to use the correct SMARTS query from the MSketch


source option.





This is


Instant JChem Version: 2.3.1 (build: 080616) JChem Version: 5.0.3 Marvin Version: 5.0.3 (build date: 2008-04-22) Java: 1.5.0_06; Java HotSpot(TM)


Client VM 1.5.0_06-b05 System: Windows XP version 5.1 running on x86; Cp1252; en_US (instantjchem) NetBeans build: 071221








Bye


Tobias

ChemAxon fa971619eb

03-10-2008 09:25:25

You are correct that there is an inconsistency in IJC.


In the 2.3.1 version the structure in the text window is always displayed in smiles format. For substructure searches this is not correct and smarts would be better. This change will be in IJC 2.4.





For instance if you use 5-methyl-1H-indole as your query structure then this is what you will see:


IJC2.3.1: CC1=CC2=C(NC=C2)C=C1


IJC2.4: [#6]C1=CC2=C(NC=C2)C=C1





The first is smiles, the second smarts.


Until 2.4 you should continue to ge the smarts from the MarvinSketch window.





Tim

User 677b9c22ff

06-10-2008 18:38:13

Thanks Tim.


Tobias