mdio 0.0.4

Basic I/O for MD trajectory files

mdio

A library of simple I/O routines for MD trajectory formats.

mdio is designed to provide basic MD file I/O capabilities. It's not supposed to replace great packages like mdtraj and mdanalysis, but may be useful when all you need is basic MD trajectory file I/O and nothing much more.

Currently mdio just supports reading and writing dcd and xtc format files.

Installation:

Easiest via pip:

% pip install mdio

Usage:

mdio provides two basic functions: xtc_open() and dcd_open(). Reading and writing is done via Frame objects that wrap the key data (coordinates, box info, timepoint).

Be aware: mdio makes no assumptions about units - it's up to the user to provide data in angstroms or nanometers, etc., as required.

Example 1: reading a dcd format file:

from mdio.dcdio import dcd_open

f = dcd_open('mydcdfile.dcd') # default is to open in "read" mode
frame = f.read_frame() # returns a frame object
coordinates = frame.crds # an [N, 3] numpy array
box = frame.box # a [3,3] numpy array
time = frame.time # time point (float) - in whatever units the dcd file used.
timestep = frame.timestep # index number of the frame (int)

while frame is not None:
    frame = f.read_frame() # returns None at the end of the file
f.close()

Example 2: writing a dcd format file:

from mdio.dcdio import dcd_open, Frame
import numpy as np

natoms = 100
crds = np.random.rand(natoms, 3)
box = np.identity(3) * 40.0
time = 1.0

myframe = Frame(crds, box=box, time=time) # crds is the only mandatory argument; box can be None and time defaults to 1.0
with dcd_open('mydcdfile.dcd', 'w') as f: # context managers supported
    f.write_frame(frame)
	

Example 3: reading an xtc format file:

Xtc format files are handled analogously to dcd format ones, but in addition it is possible to restrict reading to an initial subset of all the atoms (e.g. maybe the trajectory is of a solvated system of a total of 10,000 atoms, but the solute is just the first 200 atoms). This can increase reading performance considerably. This option is not available for dcd files as it provides no performance boost.

from mdio.xtcio import xtc_open, Frame
import numpy as np

first_atoms = 56 # Imagine we are only interested in the first 56 atoms in each snapshot
nframes = 10 # Imagine we only want the first 10 frames in the trajectory
trajectory = np.zeros((nframes, first_atoms, 3))

with xtc_open('myxtcfile.xtc') as f:
     for i in range(nframes):
	     frame = f.read_frame(natoms=first_atoms)
		 trajectory[i] = frame.crds

Author:

Charlie Laughton [e-mail]

License:

BSD 3-clause