Converting between trajectories

[lacoperon]

Hi there,

I was wondering how one would convert between box types within pytraj.

For example, I loaded in an AMBER trajectory, whose trajectory description is:

pytraj.Trajectory, 1000 frames: 
Size: 0.011064 (GB)
<Topology: 495 atoms, 495 residues, 495 mols, PBC with box type = truncoct>

Whereas when I load in a PDB file, I get a description

pytraj.Trajectory, 1 frames: 
Size: 0.000011 (GB)
<Topology: 495 atoms, 495 residues, 495 mols, non-PBC>

What exactly does the difference between PBC with box type = truncoct and non-PBC correspond to?

Note: I'm trying to concatenate the trajectories so I can project the resulting combined trajectory using eigenvalues and eigenvectors obtained from the PCA function. But, I can't do this unless I can align the PDB structure to the AMBER trajectory, which I can't do in pytraj without them being combined. Any help therein would be greatly appreciated.

Thanks

[hainm]

• How did you get the PDB file? If it does not have CRYST1 field, the box information won't show up. Below example shows that the box information is still preserved. (Basically the box define the unitcell in your explicit water simulation).

In [8]: traj = pt.load('tz2.ortho.nc', 'tz2.ortho.parm7')[:1]

In [9]: traj
Out[9]: 
pytraj.Trajectory, 1 frames: 
Size: 0.000118 (GB)
<Topology: 5293 atoms, 1704 residues, 1692 mols, PBC with box type = ortho>

In [10]: traj.save("hey.pdb")

In [12]: pt.load('hey.pdb')
Out[12]: 
pytraj.Trajectory, 1 frames: 
Size: 0.000118 (GB)
<Topology: 5293 atoms, 1704 residues, 1692 mols, PBC with box type = ortho>

• You can use traj.unitcells = an_array with shape of (n_frames, 6) and update the topology traj.top = some_topology.

Any help therein would be greatly appreciated.

I might help if you let me know your specific question here or a specific example. Cheers.

[hainm]

Another example about unitcell

In [31]: traj
Out[31]: 
pytraj.Trajectory, 1 frames: 
Size: 0.000118 (GB)
<Topology: 5293 atoms, 1704 residues, 1692 mols, non-PBC>

In [32]: traj.top.box = pt.core.Box([39.455974, 46.821517, 40.469541, 90.0, 90.0, 90.0])

In [33]: traj
Out[33]: 
pytraj.Trajectory, 1 frames: 
Size: 0.000118 (GB)
<Topology: 5293 atoms, 1704 residues, 1692 mols, PBC with box type = ortho>

[hainm]

Truncated box

In [35]: top = pt.load_topology('tz2.truncoct.parm7')

In [36]: top.box
Out[36]: <Box: truncoct, (x, y, z, alpha, beta, gamma) = (44.8903851, 44.8903851, 44.8903851, 109.471219, 109.471219, 109.471219)>

# convert to ortho-box
top.box = pt.core.Box([39.455974, 46.821517, 40.469541, 90.0, 90.0, 90.0])
In [40]: top
Out[40]: <Topology: 5827 atoms, 1882 residues, 1870 mols, PBC with box type = ortho>

[hainm]

Note: I'm trying to concatenate the trajectories so I can project the resulting combined trajectory using eigenvalues and eigenvectors obtained from the PCA function.

One of the easy ways is to combine numpy array then reassign the traj.xyz

traj.xyz = xyz # :D 

[lacoperon]

Much appreciated!!! Apologies for the delay in responding.

[hainm]

thanks.