Notes on usage#

Atom indices#

The ᴍᴏʀғᴇᴜs API uses atom indices starting from 1, so called one-based indexing. This means that the first atom in a molecule has index 1. In contrast, Python generally uses zero-based indexing for lists and arrays.

One-based indexing from xyz file#
14
PdPMe3
Pd         -0.743397 -0.000992 -1.518770                # Atom with index 1
P           1.005896  0.002102 -0.163618                # Atom with index 2
...

We believe that one-based indexing is more intuitive for the user when, e.g., getting indices visually from molecular viewers. To accomodate for this difference, atom properties in morfeus are given as dictionaries, with keys corresponding to the atom indices. These can easily be turned into regular lists and used with zero-indexing if needed:

Accessing atom properties#
>>> from morfeus import SASA, read_geometry
>>> elements, coordinates = read_geometry("Et.gjf")
>>> sasa = SASA(elements, coordinates)
>>> sasa.atom_areas
{1: 22.489947408719903,
2: 19.180448148100027,
3: 22.529947315182724,
...
>>> list(sasa.atom_areas.values())
[22.489947408719903,
19.180448148100027,
22.529947315182724,
...

Geometry files#

Many features of ᴍᴏʀғᴇᴜs makes use of atomic numbers/symbols and coordinates. These can be read from geometry files, and currently the files formats gjf (Gaussian input file) and xyz (XMOL )are supported. These files can be read with two different functions. There is also the read_geometry function that will try to guess the file type based on its suffix. If the cclib package is installed, it can be used to read many file formats.

Format

Function

gjf

read_gjf

xyz

read_xyz

*

read_cclib

Radii#

ᴍᴏʀғᴇᴜs makes extensive use of atomic radii in calculating the different descriptors. There are mainly two types of radii: vdW and covalent.

Options#

vdW

Covalent

alvarez

pyykko

bondi

crc

rahm

truhlar

  • Default vdW radii are from the curated collection in the CRC handbook

  • The only option for covalent radii at the moment is those from pyykko

  • Rahm radii are a special type derived from electron density calculation. They are the default for dispersion descriptor calculations.

Changing the type of radii is done by specifying the --radii_type keyword argument.

Sources#

The data is extracted from the excellent mendeleev package.

Without display#

Matplotlib may cause problems on computers without displays, e.g., nodes on computer clusters. This can be solved by changing the plotting backend to “agg”. This could be done at the top of the Python script using ᴍᴏʀғᴇᴜs:

import matplotlib
matplotlib.use('Agg')

Alternatively, set the environment variable in the shell before launching the script (Linux example):

export MPLBACKEND="agg"