becksteinlab / bornprofiler Goto Github PK

www.poissonboltzmann.org moved to GitHub and links such as http://www.poissonboltzmann.org/apbs/user-guide/running-apbs/input-files/elec-input-file-section/elec-keywords/dime are broken and should be fixed.

The BornProfiler documentation currently consists mostly of the doc strings or placeholders.

• Examples and use of the scripts: how should the typical user make use of the package?
  • describe workflow and individual steps
  • add figures that show typical output
• More prose is needed to explain use of the classes for developers.
• No "next" and "previous" links that allow one to quickly click through.
• Customize the theme.

@lvanderf has added major extensions (using OPM, automation, plotting) and fixes.

I created setup dirs with

apbs-bornprofile-init.py

and then want to just see the help:

(bp) oliver@spudda:~/Projects/Methods$ apbs-mem-potential.py -h
Traceback (most recent call last):
  File "/nfs/homes2/oliver/.virtualenvs/bp/bin/apbs-mem-potential.py", line 50, in <module>
    import bornprofiler.electrostatics
  File "/nfs/homes2/oliver/.virtualenvs/bp/local/lib/python2.7/site-packages/bornprofiler/electrostatics.py", line 55, in <module>
    'solvation_no_membrane': read_template('solvation_no_membrane.in'),
  File "/nfs/homes2/oliver/.virtualenvs/bp/local/lib/python2.7/site-packages/bornprofiler/bpio.py", line 382, in read_template
    fn = config.get_template(filename)
  File "/nfs/homes2/oliver/.virtualenvs/bp/local/lib/python2.7/site-packages/bornprofiler/config.py", line 252, in get_template
    templates = [_get_template(s) for s in utilities.asiterable(t)]
  File "/nfs/homes2/oliver/.virtualenvs/bp/local/lib/python2.7/site-packages/bornprofiler/config.py", line 308, in _get_template
    raise ValueError(errmsg)
ValueError: Failed to locate the template file 'solvation_no_membrane.in'.

This should work.

Replace homegrown versioning with versioneer.

On executing the example 2BG9 by using APBS-3.0.0 with given inputs I am encountering ASSERTION error. Do you have any clue whats happening here?

Volume integral = 0.00e+00 A^3
Reading kappa map data from kappa_prot_Mm.dx.gz:
0 x 0 x 0 grid
(0, 0, 0) A spacings
(0, 0, 0) A lower corner
Volume integral = 0.00e+00 A^3
Reading kappa map data from kappa_prot_Sm.dx.gz:
0 x 0 x 0 grid
(0, 0, 0) A spacings
(0, 0, 0) A lower corner
Volume integral = 0.00e+00 A^3
Reading kappa map data from kappa_cpx_Lm.dx.gz:
0 x 0 x 0 grid
(0, 0, 0) A spacings
(0, 0, 0) A lower corner
Volume integral = 0.00e+00 A^3
Reading kappa map data from kappa_cpx_Mm.dx.gz:
0 x 0 x 0 grid
(0, 0, 0) A spacings
(0, 0, 0) A lower corner
Volume integral = 0.00e+00 A^3
Reading kappa map data from kappa_cpx_Sm.dx.gz:
0 x 0 x 0 grid
(0, 0, 0) A spacings
(0, 0, 0) A lower corner
Volume integral = 0.00e+00 A^3
Got paths for 6 charge maps
Reading charge map data from charge_prot_Lm.dx.gz:
0 x 0 x 0 grid
(0, 0, 0) A spVgrid_value: Error -- no data available!
VASSERT: ASSERTION FAILURE! filename /home/ubuntu/git/apbs-pdb2pqr/apbs/src/mg/vgrid.c, line 193, (0)
-- job complete: results in job_0000.out

We should print a message when a script runs (log at INFO), e.g

bornprofiler VERSION (https://github.com/Becksteinlab/BornProfiler)
Copyright (c) ...
Released under the GNU Public Licence, version 3.

The current approach to use a weighted histogram to build the PMF on a grid can lead to 3D PMFs that are patchy. Instead we should just use a KDE to estimate the density from the compute points and then sample the KDE on an arbitrary resolution grid to write a DX file.

• scipy Gaussian KDE

If apbs was compiled without zlib compression --disable-zlib, i.e., without support for reading/writing gz compressed files, then drawmembrane will fail

ERROR: dielxS.dx.gz not found in current directory.
       See -h for usage.
bornprofiler.electrostatics: CRITICAL drawmembrane failed [returncode 1]
Traceback (most recent call last):
  File "$HOME/anaconda3/envs/bp/bin/apbs-mem-potential.py", line 7, in <module>
    exec(compile(f.read(), __file__, 'exec'))
  File "$HOME/APBS/BornProfile/scripts/apbs-mem-potential.py", line 109, in <module>
    A.generate()
  File "$HOME/APBS/BornProfile/bornprofiler/electrostatics.py", line 372, in generate
    self.run_drawmembrane()
  File "$HOME/APBS/BornProfile/bornprofiler/electrostatics.py", line 198, in run_drawmembrane
    raise EnvironmentError(errno.ENODATA, self.drawmembrane, errmsg)
EnvironmentError: [Errno 96] draw_membrane2a: 'drawmembrane failed [returncode 1]'

The current GLIC example contains pdb and cfg as input.
In order to launch the whole BornProfiler workflow, a pqr and a grid dat file are required.

Creating DX files for PMF grids fails because BP did not keep up with GridDataFormats.

Code needs to be changed to work with latest GDF.

Reported by @ChenouZhang