6.8. GRO file format — MDAnalysis.coordinates.GRO

Classes to read and write Gromacs GRO coordinate files; see the notes on the GRO format which includes a conversion routine for the box.

6.8.1. Writing GRO files

By default any written GRO files will renumber the atom ids to move sequentially from 1. This can be disabled, and instead the original atom ids kept, by using the reindex=False keyword argument. This is useful when writing a subsection of a larger Universe while wanting to preserve the original identities of atoms.

For example:

>>> u = mda.Universe()`

>>> u.atoms.write('out.gro', reindex=False)

# OR
>>> with mda.Writer('out.gro', reindex=False) as w:
...     w.write(u.atoms)

6.8.2. Classes

class MDAnalysis.coordinates.GRO.Timestep(n_atoms, **kwargs)[source]

Create a Timestep, representing a frame of a trajectory

  • n_atoms (int) – The total number of atoms this Timestep describes
  • positions (bool, optional) – Whether this Timestep has position information [True]
  • velocities (bool (optional)) – Whether this Timestep has velocity information [False]
  • forces (bool (optional)) – Whether this Timestep has force information [False]
  • reader (Reader (optional)) – A weak reference to the owning Reader. Used for when attributes require trajectory manipulation (e.g. dt)
  • dt (float (optional)) – The time difference between frames (ps). If time is set, then dt will be ignored.
  • time_offset (float (optional)) – The starting time from which to calculate time (in ps)

Changed in version 0.11.0: Added keywords for positions, velocities and forces. Can add and remove position/velocity/force information by using the has_* attribute.


unitcell dimensions (A, B, C, alpha, beta, gamma)


8.00170   8.00170   5.65806   0.00000   0.00000   0.00000   0.00000   4.00085   4.00085


CRYST1   80.017   80.017   80.017  60.00  60.00  90.00 P 1           1

XTC: c.trajectory.ts._unitcell:

array([[ 80.00515747,   0.        ,   0.        ],
       [  0.        ,  80.00515747,   0.        ],
       [ 40.00257874,  40.00257874,  56.57218552]], dtype=float32)
class MDAnalysis.coordinates.GRO.GROReader(filename, convert_units=None, n_atoms=None, **kwargs)[source]

Reader for the Gromacs GRO structure format.

Changed in version 0.11.0: Frames now 0-based instead of 1-based

Writer(filename, n_atoms=None, **kwargs)[source]

Returns a CRDWriter for filename.

Parameters:filename (str) – filename of the output GRO file
Return type:GROWriter
class MDAnalysis.coordinates.GRO.GROWriter(filename, convert_units=None, n_atoms=None, **kwargs)[source]

GRO Writer that conforms to the Trajectory API.

Will attempt to write the following information from the topology:
  • atom name (defaults to ‘X’)
  • resnames (defaults to ‘UNK’)
  • resids (defaults to ‘1’)


The precision is hard coded to three decimal places

Changed in version 0.11.0: Frames now 0-based instead of 1-based

Changed in version 0.13.0: Now strictly writes positions with 3dp precision. and velocities with 4dp. Removed the convert_dimensions_to_unitcell method, use Timestep.triclinic_dimensions instead. Now now writes velocities where possible.

Changed in version 0.18.0: Added reindex keyword argument to allow original atom ids to be kept.

Set up a GROWriter with a precision of 3 decimal places.

  • filename (str) – output filename
  • n_atoms (int (optional)) – number of atoms
  • convert_units (str (optional)) – units are converted to the MDAnalysis base format; None selects the value of MDAnalysis.core.flags [‘convert_lengths’]
  • reindex (bool (optional)) – By default, all the atoms were reindexed to have a atom id starting from 1. [True] However, this behaviour can be turned off by specifying reindex =False.


To use the reindex keyword, user can follow the two examples given below.:

u = mda.Universe()

Usage 1:

u.atoms.write('out.gro', reindex=False)

Usage 2:

with mda.Writer('out.gro', reindex=False) as w:
fmt = {'box_orthorhombic': '{box[0]:10.5f}{box[1]:10.5f}{box[2]:10.5f}\n', 'box_triclinic': '{box[0]:10.5f}{box[4]:10.5f}{box[8]:10.5f}{box[1]:10.5f}{box[2]:10.5f}{box[3]:10.5f}{box[5]:10.5f}{box[6]:10.5f}{box[7]:10.5f}\n', 'n_atoms': '{0:5d}\n', 'xyz': '{resid:>5d}{resname:<5.5s}{name:>5.5s}{index:>5d}{pos[0]:8.3f}{pos[1]:8.3f}{pos[2]:8.3f}\n', 'xyz_v': '{resid:>5d}{resname:<5.5s}{name:>5.5s}{index:>5d}{pos[0]:8.3f}{pos[1]:8.3f}{pos[2]:8.3f}{vel[0]:8.4f}{vel[1]:8.4f}{vel[2]:8.4f}\n'}

format strings for the GRO file (all include newline); precision of 3 decimal places is hard-coded here.


Write selection at current trajectory frame to file.

Parameters:obj (AtomGroup or Universe or Timestep) –


The GRO format only allows 5 digits for resid and atom number. If these numbers become larger than 99,999 then this routine will chop off the leading digits.

Changed in version 0.7.6: resName and atomName are truncated to a maximum of 5 characters

Changed in version 0.16.0: frame kwarg has been removed

6.8.3. Developer notes: GROWriter format strings

The GROWriter class has a GROWriter.fmt attribute, which is a dictionary of different strings for writing lines in .gro files. These are as follows:


For the first line of the gro file, supply the number of atoms in the system. E.g.:


An atom line without velocities. Requires that the ‘resid’, ‘resname’, ‘name’, ‘index’ and ‘pos’ keys be supplied. E.g.:

fmt['xyz'].format(resid=1, resname='SOL', name='OW2', index=2, pos=(0.0, 1.0, 2.0))
As above, but with velocities. Needs an additional keyword ‘vel’.

The final line of the gro file which gives box dimensions. Requires the box keyword to be given, which should be the three cartesian dimensions. E.g.:

fmt['box_orthorhombic'].format(box=(10.0, 10.0, 10.0))
As above, but for a non orthorhombic box. Requires the box keyword, but this time as a length 9 vector. This is a flattened version of the (3,3) triclinic vector representation of the unit cell. The rearrangement into the odd gromacs order is done automatically.