Clarification on VFLP input

VFLP VFLP 1
#1

So in the input-files directory, I have one massive .smi file, where each line has a SMILES string for each compound. So then in the all.ctrl file I specified the collection_folder to be …/input-files. However, when I ran VFLP I got no output. Is the problem the collection_folder parameter that I specified? Or is it the way my .smi file has been formatted (needing one file per SMILES string vs having them all in one file)?

Thanks

#2

Dear @janujan9954 ,

you need one SMILES file per ligand.

This was already mentioned in more detail in other threads, such as :

In summary, the ligand library of VFVS and VFLP are of the same structure (one file per ligand, with meta tranches, tranches and collections). Except that for VFLP it is a SMILES file instead of a PDBQT or MOL2 file.

Hope this helps,
Christoph

1 Like
#3

Hi Cristoph,

Thanks for the prompt response. I am a little confused however; in VFVS the ligands are stored in tar.gz files, which contain tar.gz files in those as well. If the ligand library of VFVS and VFLP are similar, do I need to compress the directory that holds all of the .smi files into a tar.gz file, and specify that in my collection_folder parameter?

Thanks

Thanks

#4

Just to clarify, if I want to convert 20 million compounds, I will need 20 million separate .smi files all organized similarly to the structure for VFVS? I feel like I am misunderstanding something…

#5

Hi @janujan9954 ,

Yes, each ligand is stored in its own SMILES file.

Yes, the tar archives and compressed tar archives are needed for VFLP as well.

Hope this helps,
Christoph