diff extract_rmsd.py @ 4:4c36f5ad2799 draft

"planemo upload for repository https://github.com/galaxycomputationalchemistry/galaxy-tools-compchem/ commit 1b23e024af45cc0999d9142d07de6897d4189ec2"
author chemteam
date Mon, 24 Aug 2020 16:53:30 -0400
parents
children dfda5e713926
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/extract_rmsd.py	Mon Aug 24 16:53:30 2020 -0400
@@ -0,0 +1,129 @@
+import argparse
+import json
+
+import MDAnalysis as m
+from MDAnalysis.analysis import align, rms
+from MDAnalysis.analysis.base import AnalysisFromFunction
+from MDAnalysis.coordinates.memory import MemoryReader
+
+import numpy as np
+
+
+def calc_rmsd(str_files, traj_files, ref_str, str_format, traj_format,
+              ref_str_format, filepath_out, group, start, end, step,
+              fitting_atoms):
+    """
+    the function will cycle through range 0 to no_t and load all files found.
+
+    str_files: text file with filepaths for structures, one on each line
+    traj_files: text file with filepaths for trajectories, one on each line
+    ref_str: reference structure for fitting
+    filepath_in: directory where the files are located
+    filepath_out: pickle file where results (3D matrix) should be saved to
+
+    group: atoms for which RMSD should be calculated;
+        use the MDAnalysis selection language
+    fitting_atoms: atoms used for str alignment prior to RMSD calculation;
+        use the MDAnalysis selection language
+
+    start: first trajectory frame to calculate RMSD
+    end: last trajectory frame to calculate RMSD
+    step: how frequently frames are sampled between start and end; obviously,
+        the larger the step, the quicker the script finishes
+    """
+
+    # open list of files
+    with open(str_files) as f1, open(traj_files) as f2:
+        str_file_list = f1.read().strip().split('\n')
+        traj_file_list = f2.read().strip().split('\n')
+
+        if sum(1 for line in f1) != sum(1 for line in f2):
+            raise IOError('Number of structure and trajectory files unequal.')
+
+    no_t = len(traj_file_list)
+
+    data = np.zeros((no_t, no_t,
+                    int((end - start)/step + ((end - start) % step > 0))))
+
+    # load files
+    universes = {}
+
+    for traj in range(no_t):
+        mobile = m.Universe(str_file_list[traj], traj_file_list[traj],
+                            format=traj_format, topology_format=str_format)
+        ref = m.Universe(ref_str, topology_format=ref_str_format)
+
+        mobile.trajectory[-1]  # set mobile trajectory to last frame
+        ref.trajectory[0]  # set reference trajectory to first frame
+
+        # perform alignment
+        align.AlignTraj(mobile, ref, select=fitting_atoms,
+                        in_memory=True).run()
+
+        grp = mobile.select_atoms(group)
+        universes[traj] = m.core.universe.Merge(grp)  # create Universe w grp
+        coordinates = AnalysisFromFunction(lambda ag: ag.positions.copy(),
+                                           grp).run().results  # write to uv
+        universes[traj].load_new(coordinates, format=MemoryReader)
+
+    print("All trajs loaded by MDAnalysis")
+
+    # calculate differences
+    for traj1 in range(no_t):
+        print("Calculating differences for traj {}".format(traj1))
+        for traj2 in range(traj1):
+
+            u1 = universes[traj1]
+            u2 = universes[traj2]
+
+            l1 = u1.select_atoms(group)
+            l2 = u2.select_atoms(group)
+
+            rmsd = rms.RMSD(l1, l2)
+
+            rmsd.run()
+
+            data[traj1, traj2] = rmsd.rmsd[:, 2]
+            data[traj2, traj1] = rmsd.rmsd[:, 2]
+
+    with open(filepath_out, 'w') as f:
+        json.dump(data.tolist(), f, indent=4, sort_keys=True)
+
+    print("Done!")
+    return
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--trajs', required=True,
+                        help='File containing trajectory filepaths.')
+    parser.add_argument("--strs",
+                        help='File containing structure filepaths.')
+    parser.add_argument("--ref-str",
+                        help='File containing reference structure.')
+    parser.add_argument('--traj-format', required=True,
+                        help='Trajectory format.')
+    parser.add_argument("--str-format", help='Structure format.')
+    parser.add_argument("--ref-str-format",
+                        help='Reference structure format.')
+    parser.add_argument('-o', '--outfile',
+                        help="Path to the output JSON file")
+    parser.add_argument('--group', help="Atoms for which RMSD should be"
+                        "calculated in MDAnalysis selection language")
+    parser.add_argument('--fitting', help="Fitting atoms for alignment"
+                        "prior to RMSD calculation")
+    parser.add_argument('--start', type=int,
+                        help="First trajectory frame to calculate RMSD")
+    parser.add_argument('--end', type=int,
+                        help="Last trajectory frame to calculate RMSD")
+    parser.add_argument('--step', type=int,
+                        help="Frame sampling frequency for RMSD calculation")
+    args = parser.parse_args()
+
+    calc_rmsd(args.strs, args.trajs, args.ref_str, args.str_format,
+              args.traj_format, args.ref_str_format, args.outfile,
+              args.group, args.start, args.end, args.step, args.fitting)
+
+
+if __name__ == "__main__":
+    main()