Pavel

This would be a valuable fix for the script. However, this vina_dock.py script is obsolete and it will be removed in some future revisions of the repository. Now I suggest...

Hi, @AnzorGozalishvili, could you attach your input files to reproduce this behavior? It seems like a simple issue.

It is assumed that these are conformations in which those compounds should bind to the target, but it is not 100% necessary. Doing docking you may use this conformation as...

From which branch did you install psearch? You have to use `gen_pharms` branch, it is the most recent. Unfortunately we still did not fix all remaining bugs and merge it...

Does it return any pharmacophore model? If not, it may happen that it cannot create training sets. I got something similar in the past and it would be reasonable to...

Fixes: 1. Reduce the chunk size in gen_db to 1. This will not result in much overhead but will distribute calculations better, especially small data sets on multi-cpu machines. 2....

You have to generate a DB with conformers of your molecules and apply `screen_db` ( utility provided with the module). Examples are provided in README. For example: `screen_db -d DB.dat...

You are right, f7 is expected to designate that there are 7 features in a pharmacophore. This seems like a bug, but after quick investigation I could not figure out...

> I would like to understand what the criteria are for selecting the best model since there is no alignment. Recall> precision> FPR, etc ? if you ask about selection...

There is no classical feature radius. This alignment-free approach is somewhat different from others. A some sort of radius is a binning step (`bin_step`), which can be specified during model...