Hi all -
I am new to Knime, and am generally very happy with it. I am mostly using it to prepare libraries of compounds for virtual screening: taking SMILES strings of millions of compounds, convert to 3D representations, minimize, Lipinski Filtering, etc., and then finally writing to mol2 format with the “mol2 writer” (in the chemistry I/O node collection).
My initial SMILES file contains the structure (“col0”) and a catalog number (aka title; as “col1”) for each molecule, which is read initially and passed through each node. I confirm that “col1” is passed through each node with the MarvinView tool.
From what I can tell, the mol2 writer node does not seem to have functionality to write titles, which is unfortunate for virtual screening, since the initial catalog number is very useful However, the SDF writer is more robust, allowing writing of titles. (But mol2 is preferred for virtual screening, since it supports partial atomic charges)
Currently if I want to encode the mol2 with a title data, I need to write the mol2 file and then list of titles (“col1”) as a CSV file, and merge them with an outside script, which is tedious.
I guess I am wondering if the mol2 writer can be modified to act like the SDF writer, and allow incorporation of the title? I think this would improve the usefulness of the mol2 writer. Thanks!!