Error opening SDF file with MarvinView or MolImporter

Hi,

when I try to open the attached file with MarvinView I get the following stack trace -

chemaxon.formats.MolFormatException: Error parsing SMILES string '2|LigandData|sdf|2|dock1' at character 1 ('2')
 at chemaxon.marvin.io.formats.smiles.SmilesImport.readMol0(Unknown Source)
 at chemaxon.marvin.io.formats.smiles.SmilesImport.readMol(Unknown Source)
 at chemaxon.marvin.io.formats.smiles.CxsmilesImport.readMol(Unknown Source)
 at chemaxon.marvin.io.formats.smiles.SmilesImport.readMol(Unknown Source)
 at chemaxon.marvin.io.MRecordImporter.readStructure(Unknown Source)
 at chemaxon.marvin.io.MRecordImporter.readDoc(Unknown Source)
 at chemaxon.marvin.io.MRecordImporter.readDoc(Unknown Source)
 at chemaxon.marvin.io.MRecordImporter.readDoc0(Unknown Source)
 at chemaxon.marvin.io.MRecordImporter.readDoc(Unknown Source)
 at chemaxon.formats.MolImporter.readDoc(Unknown Source)
 at chemaxon.formats.MolImporter.nextDoc(Unknown Source)
 at chemaxon.marvin.view.MDocStorage.readDoc(Unknown Source)
 at chemaxon.marvin.view.MDocStorage.tryToExtend(Unknown Source)
 at chemaxon.marvin.view.MDocStorage.getMainDoc(Unknown Source)
 at chemaxon.marvin.view.MDocStorage.getMainDoc(Unknown Source)
 at chemaxon.marvin.view.swing.TableSupport$11.run(Unknown Source)
 at chemaxon.marvin.view.swing.TableSupport.setDocSource(Unknown Source)
 at chemaxon.marvin.view.swing.TableSupport.access$600(Unknown Source)
 at chemaxon.marvin.view.swing.TableSupport$UpdateTask.run0(Unknown Source)
 at chemaxon.marvin.view.swing.TableSupport$UpdateTask.run(Unknown Source)
 at chemaxon.marvin.view.SequentialScheduler$CmpRunnable.run(Unknown Source)
 at chemaxon.marvin.view.SequentialScheduler$1.run(Unknown Source)

If I try to read the sdf file with

File results = new File(m_Output.getStringValue());
FileInputStream fileInputStream = new FileInputStream(results);
MolImporter molImporter = new MolImporter(fileInputStream);
while ((molecule = molImporter.read()) != null){
}

I get the following error -

Error parsing SMILES string '2|LigandData|sdf|2|dock1' at character 1 ('2')

The SDF file looks to be a standard file and if i delete the first line MarvinView will read it ok and I can read it using MolImporter. The file is generated by another application on a Linux system and I am working on it from Windows (after running unix2dos on it , but get the same result even if I don't run unix2dos).

Can you suggest why I am getting these errors (and a fix if possible)?

Thanks

Best regards

Bob

Hi Bob,

this is because of the pipe(|) characters, the SMILES recognizer recognizes the file, and Marvin try to import it as a smiles. You can use the MolImporter(String, String) constructor, and add "sdf" as a second parameter, this will explicitly tell marvin to import the file as an sdf.

In Marvin 5.6 the recognition will be imporved, but this is a hard one, and the solution is also not perfect, but much more effective (file format recognition is a hard thing to make it perfect, and has contionious improvments in marvin, as we discover such issues), there will be the following rule in 5.6:
If the first line contains two pipe character, and the first one has a white space before, and the second one has a white space, or an eof after then the file is considered to be in cxsmiles format.

Current marvin versions has the following rule: if the first line contains a pipe character, and not starting with the "Vj" character sequence, then the file is considered to be in cxsmiles format. This rule causing the problem.

Regards,
Istvan

Thanks for your help and explanation Istvan.

I will try MolImporter(String, String) and look forward to the improvements in Marvin 5.6.

Cheers

Bob