Supported file formats


Most commmon usage: load a single file with all frame

# read amber format
In [1]: import pytraj as pt

In [2]: traj = pt.iterload('', 'tz2.ortho.parm7')

In [3]: traj
pytraj.TrajectoryIterator, 10 frames: 
Size: 0.001183 (GB)
<Topology: 5293 atoms, 1704 residues, 1692 mols, PBC with box type = ortho>

Most commmon usage: load a single file with frame stride

# load only frame 0, 2, 4, 6 (python use 0-based index and skip final index)
traj = pt.iterload('', 'tz2.parm7', frame_slice=[(0, 8, 2),])

Load many files at once

In [1]: ls remd.x.*
remd.x.000  remd.x.003  remd.x.006  remd.x.009  remd.x.012  remd.x.015
remd.x.001  remd.x.004  remd.x.007  remd.x.010  remd.x.013  remd.x.016
remd.x.002  remd.x.005  remd.x.008  remd.x.011  remd.x.014  remd.x.017

In [2]: traj = pt.iterload('remd.x.*', 'myparm.parm7')

In [3]: traj
<pytraj.TrajectoryIterator, 40000 frames, include:
<Topology: 17443 atoms, 5666 residues, 5634 mols, PBC with box type = truncoct>>

In [4]: traj._estimated_MB
Out[4]: 15969.54345703125

Load many files with frame stride

Example: load 1000 frames from two trajectories (500 each), skip every two frames.

In [1]: traj = pt.iterload(['remd.x.000', 'remd.x.001'], 'myparm.parm7', frame_slice=[(0, 1000, 2), (0, 1000, 2)])

In [2]: traj
<pytraj.TrajectoryIterator, 1000 frames, include:
<Topology: 17443 atoms, 5666 residues, 5634 mols, PBC with box type = truncoct>>

which is similiar to cpptraj input:

parm myparm.parm7
trajin remd.x.000 1 1000 2
trajin remd.x.001 1 1000 2


cpptraj uses 1-based index for input while pytraj used 0-based index.


# write to CHARMM dcd format
In [4]: pt.write_traj('test.dcd', traj, overwrite=True)

# write with given frames
In [5]: pt.write_traj('test2.dcd', traj, frame_indices=[0, 3, 7, 9], overwrite=True)

Supported file formats

pytraj.write_traj(filename, traj, format='infer', top=None, frame_indices=None, overwrite=False, force=False, velocity=False, options='')

write Trajectory-like or iterable object to trajectory file


filename : str

traj : Trajectory-like or iterator that produces Frame or 3D ndarray with shape=(n_frames, n_atoms, 3)

format : str, default ‘infer’

if ‘inter’, detect format based on extension. If can not detect, use amber mdcdf format.

top : Topology, optional, default: None

frame_indices: array-like or iterator that produces integer, default: None

If not None, only write output for given frame indices

overwrite: bool, default: False

velocity : bool, default False

if True, write velocity. Make sure your trajectory or Frame does have velocity

force : bool, default False

if True, write force. Make sure your trajectory or Frame does have force

options : str, additional cpptraj keywords


Format Extension
Amber Trajectory .crd
Amber NetCDF .nc
Amber Restart .rst7
Amber NetCDF .ncrst
Charmm DCD .dcd
PDB .pdb
Mol2 .mol2
Scripps .binpos
Gromacs .trr
SQM Input .sqm

‘options’ for writing to pdb format (cptraj manual):

dumpq:       Write atom charge/GB radius in occupancy/B-factor columns (PQR format)."
parse:       Write atom charge/PARSE radius in occupancy/B-factor columns (PQR format)."
vdw:         Write atom charge/VDW radius in occupancy/B-factor columns (PQR format)."
pdbres:      Use PDB V3 residue names."
pdbatom:     Use PDB V3 atom names."
pdbv3:       Use PDB V3 residue/atom names."
teradvance:  Increment record (atom) # for TER records (default no)."
terbyres:    Print TER cards based on residue sequence instead of molecules."
model:       Write to single file separated by MODEL records."
multi:       Write each frame to separate files."
chainid <c>: Write character 'c' in chain ID column."
sg <group>:  Space group for CRYST1 record, only if box coordinates written."
include_ep:  Include extra points."
conect:      Write CONECT records using bond information.");


>>> import pytraj as pt
>>> traj = pt.datafiles.load_tz2_ortho()
>>> pt.write_traj("output/", traj, overwrite=True) # write to amber netcdf file
>>> # write to multi pdb files (t.pdb.1, t.pdb.2, ...)
>>> pt.write_traj("output/t.pdb", traj, overwrite=True, options='multi')
>>> # write all frames to single pdb file and each frame is seperated by "MODEL" word
>>> pt.write_traj("output/t.pdb", traj, overwrite=True, options='model')
>>> # write to DCD file
>>> pt.write_traj("output/test.dcd", traj, overwrite=True)
>>> # write to netcdf file from 3D numpy array, need to provide Topology
>>> xyz =
>>> top =
>>> pt.write_traj("output/", xyz,, overwrite=True)
>>> pt.write_traj("output/", xyz,, overwrite=True)
>>> # you can make a fake Topology to write xyz coordinates too
>>> n_atoms = xyz.shape[1]
>>> top2 =
>>> pt.write_traj("output/", xyz, top=top2, overwrite=True)

‘options’ for writing to amber netcdf format (cptraj manual):

remdtraj: Write temperature to trajectory (makes REMD trajectory)."
velocity: Write velocities to trajectory."
force: Write forces to trajectory.");

‘options’ for writing to amber netcdf restart format(cptraj manual):

novelocity: Do not write velocities to restart file."
notime:     Do not write time to restart file."
remdtraj:   Write temperature to restart file."
time0:      Time for first frame (default 1.0)."
dt:         Time step for subsequent frames, t=(time0+frame)*dt; (default 1.0)");
keepext     Keep filename extension; write '<name>.<num>.<ext>' instead (example: myfile.1.rst7)

‘options’ for writing to mol2 format (cptraj manual):

single   : Write to a single file."
multi    : Write each frame to a separate file."
sybyltype: Convert Amber atom types (if present) to SYBYL types.");

‘options’ for other formats:

please check