Re: [Rdkit-discuss] Request for Assistance: Understanding InChI to Mol Conversion Issue in RDKit

[S J. S. <swa...@gm...>]

Perhaps provide some examples were this failure happens.

Sent from Gmail Mobile


On Tue, Nov 28, 2023 at 7:35 PM 李大舟 <lid...@sy...> wrote:

> Dear RDKit Developers and Maintainers,
>
> I hope this email finds you well. My name is Dr. Dazhou Li, and I am a
> researcher working on the development of a tool for extracting chemical
> compound structures recognized by OCR (Optical Character Recognition)
> technology. I have been using the RDKit library for a crucial step in this
> process, specifically the rdkit.Chem.inchi.MolFromInchi() function, to
> convert InChI-format strings into Mol format representations.
>
> Firstly, I would like to express my gratitude for the excellent work you
> have done in developing and maintaining the RDKit library, which has been
> an invaluable resource in my research. The library has consistently
> delivered high-quality results in various aspects of chemical informatics,
> and I appreciate your dedication to its development.
>
> However, I have encountered a specific issue with the
> rdkit.Chem.inchi.MolFromInchi() function that I hope you can help me
> understand and resolve. When attempting to convert InChI-format strings
> generated by my tool, some of them fail with an error message reporting
> "NaN." Since the rdkit.Chem.inchi.MolFromInchi() function calls C++ code, I
> am unable to directly inspect its execution or source code to diagnose the
> issue.
>
> My primary request is for assistance in understanding the internal
> workings of the rdkit.Chem.inchi.MolFromInchi() function, specifically the
> checking process or generation step that leads to the "NaN" error when
> certain InChI-format strings are processed. It is crucial for my research
> to determine at which point in the execution of this function my generated
> InChI-formatted strings are considered unreasonable, as this information
> will help me refine my tool's output to be compatible with RDKit.
>
> I understand that the RDKit library is a complex and comprehensive
> toolkit, and I appreciate the complexity involved in diagnosing such
> issues. However, any insights or guidance you can provide regarding the
> problematic cases and the internal processes of the
> rdkit.Chem.inchi.MolFromInchi() function would be immensely valuable to me
> and would help me ensure the compatibility of my tool with RDKit.
>
> If possible, I would be grateful for access to relevant documentation or
> insights into the specific error conditions that may lead to the "NaN"
> result. Additionally, any suggestions or best practices for generating
> InChI-format strings that are more likely to be successfully processed by
> RDKit would be greatly appreciated.
>
> Thank you for your time and consideration. I look forward to your response
> and hope that we can collaborate to resolve this issue and enhance the
> compatibility of my tool with the RDKit library.
>
> Please feel free to reach out to me if you require any additional
> information or if there are specific details about my tool or the
> InChI-format strings that would aid in diagnosing the issue.
>
> Best regards,
>
> Dr. Dazhou Li
> Shenyang University of Chemical Technology
> _______________________________________________
> Rdkit-discuss mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>