Extended SMILES import

User 677b9c22ff

24-09-2007 21:19:51

Hi,


I get some errors with SMILES which work perfectly in Mview.


[H]n1cnc2c(ncnc12)C(=N)NN


[H]n1cnc2ncnc(NC(=O)OC)c12


I think IJChem thinks that they are SMARTS?


Would it be possible to allow ChemAxon extended SMILES during import?


Tobias





Error: Failed to import row 4,506


org.springframework.jdbc.UncategorizedSQLException: java.io.IOException: file format not recognized; uncategorized SQLException for SQL []; SQL state [null]; error code [0]; java.io.IOException: file format not recognized; nested exception is java.sql.SQLException: java.io.IOException: file format not recognized


Error: Failed to import row 4,705


org.springframework.jdbc.UncategorizedSQLException: java.io.IOException: file format not recognized; uncategorized SQLException for SQL []; SQL state [null]; error code [0]; java.io.IOException: file format not recognized; nested exception is java.sql.SQLException: java.io.IOException: file format not recognized

ChemAxon fa971619eb

25-09-2007 08:41:10

Sorry, I can't reproduce this problem.


IJC does import CXSMILES format, and those 2 smiles strings you specified import fine for me in IJC 2.0.2





Can you provide further info about this issue?





Tim

User 677b9c22ff

26-09-2007 06:02:16

Hi,


I think you need to enter some SMILES before and then enter some SMARTS that will bring the import error.


Tobias

ChemAxon fa971619eb

27-09-2007 14:45:18

Well, those 2 structures you mention are not smarts, just normal smiles.





If you do have mixed smiles and smarts in a file you can get problems if the first part of the file just contains smiles. This is because IJC tries to guess the best import settings by reading the begining of the file. The exact strategy is a little complex, but basically it reads the file until it finds no changes for 100 records, so a file where you get smiles at the start and then a smarts structure (or a reaction...) after a large number of normal smiles can give problems.





The problem that will happen is that IJC tries to use the most specific JChem table type for the data. When the data only contains discrete structures (e.g smiles) it will use Molecules as the table type. This table type is for discrete structures and does not allow query structures (e.g. smarts) or reactions. So when you come across a query structure later in the file you will see errors like this:





Error: Failed to import row 9,998


org.springframework.jdbc.UncategorizedSQLException: Inserting a query or Markush structure is not allowed for table: "APP.SMILES_THEN_SMARTS"; uncategorized SQLException for SQL []; SQL state [null]; error code [0];





The solution is either to set the table content manually to Any Structures, or to parse the complete file at the start of the import process (set the number of records to zero and press the 'Read more' button). This will give IJC its best chance to determine the best settings.





Tim

User 677b9c22ff

27-09-2007 19:10:03

Hi Tim,


thank you. I will do that. It just reminds me to check every time the number of compounds and structures etc. during import and export.


Tobias