patonlab / dbstep Goto Github PK

This can be done by making sure there is no longer a reference to it. Currently, once it is done being used, the reference is not set to None, so it just continues to unnecessarily exist in memory. The variable name in dbstep.py is self.unocc_grid.

DBSTEP atom references are 1-indexed (starting the atom counts from 1, 2, 3..) and atom "0" should not be a valid atom index.

Guilian can you update the sections for which I've added headers to the README? Also would be nice to show some graphics of sterimol and vectorised sterimol from pymol as well!

When I try to use the python API to find the buried volume for a given molecule, I get the error TypeError: Unexpected keyword argument {'n_jobs': -1}. Below is a sample code to show how I get this error.
To install dbstep, I used conda install -c conda-forge dbstep and it installed version 1.0.0.

import urllib.request
import dbstep.Dbstep as db
url = "https://raw.githubusercontent.com/patonlab/DBSTEP/master/dbstep/examples/1Nap.xyz"
filename, headers = urllib.request.urlretrieve(url, filename="1Nap.xyz")
db.dbstep("1Nap.xyz", atom1=2, volume=True).bur_vol

Currently, we convert the list of indices to a set and then back to a list to remove duplicates. This does not scale very well compared to a simple alternative: make a bit mask (each bit marking whether a point is occupied or not) and, as you are scanning out the sphere's, set the needed indices to true. This means there will be no duplicates as each value is restricted to true or false. The overall grid can then be indexed using this mask to make the occupied grid.

Just wanted to enquire about a potential update to the DBSTEP PyPi and conda. The versions listed are 1.0 and still use the old scipy variable “n_jobs” instead of “workers”, unlike the latest version on github (DBSTEP 1.1.0).
Many thanks and apologies if you are already aware of this.

When excute the following command line:
python -m dbstep examples/smiles.txt --2d --fg "C(O)=O" --maxpath 5 --2d-type mcgowan
the error appears:
Traceback (most recent call last):
File "C:\Users\DELL\anaconda3\lib\runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
File "C:\Users\DELL\anaconda3\lib\runpy.py", line 87, in run_code
exec(code, run_globals)
File "C:\Users\DELL\anaconda3\lib\site-packages\dbstep_main.py", line 17, in
sys.exit(Dbstep.main())
File "C:\Users\DELL\anaconda3\lib\site-packages\dbstep\Dbstep.py", line 531, in main
vec_df = graph.mol_to_vec(file,options.shared_fg,options.voltype,options.max_path_length,options.verbose)
File "C:\Users\DELL\anaconda3\lib\site-packages\dbstep\graph.py", line 106, in mol_to_vec
mol,prop = make_mol_obj(line)
File "C:\Users\DELL\anaconda3\lib\site-packages\dbstep\graph.py", line 75, in make_mol_obj
return mol,prop
UnboundLocalError: local variable 'prop' referenced before assignment

For %Vbur calculations, memory use and speed could be improved by only generating points for atoms which intersect with the probe sphere. This should be determinable by a simple calculation: if the distance of their center points is less than the sum of their radii.

This should be determined before fitting the grid, and should generally allow for much smaller grids to be required (depending on the radii you are probing).