def calc_deshaw_centroid_alpha_cartesian(ptlist=None):
""" Calc RMSD from list of trajectories
"""
if ptlist is None:
pts = deshaw.loadpts(skip=100, filt=deshaw.FILTER['alpha'])
else:
pts = ptlist
# sums = np.zeros(shape=(5, 58, 3))
# cnts = [0 for i in range(5)]
groupby = [[] for i in range(5)]
label = deshaw.loadlabels_aslist()
for idx, pt in enumerate(pts):
# idx = math.floor(i/10)
try:
state = label[idx]
if state == label[idx - 2] == label[idx - 1] == label[
idx + 1] == label[idx + 2]:
# sums[state] += pt
# cnts[state] += 1
groupby[state].append(pt)
except IndexError as err:
pass # ignore idx errors due to tail end of DEShaw data
# cent = [sums[i] / cnts[i] for i in range(5)]
cent = np.zeros(shape=(5, 58, 3))
for i in range(5):
cent[i] = ndimage.measurements.center_of_mass(np.array(groupby[i]))
return np.array(cent)
def calc_deshaw_centroid_alpha_cartesian(ptlist=None):
""" Calc RMSD from list of trajectories
"""
if ptlist is None:
pts = deshaw.loadpts(skip=100, filt=deshaw.FILTER['alpha'])
else:
pts = ptlist
# sums = np.zeros(shape=(5, 58, 3))
# cnts = [0 for i in range(5)]
groupby = [[] for i in range(5)]
label = deshaw.loadlabels_aslist()
for idx, pt in enumerate(pts):
# idx = math.floor(i/10)
try:
state = label[idx]
if state == label[idx-2] == label[idx-1] == label[idx+1] == label[idx+2]:
# sums[state] += pt
# cnts[state] += 1
groupby[state].append(pt)
except IndexError as err:
pass # ignore idx errors due to tail end of DEShaw data
# cent = [sums[i] / cnts[i] for i in range(5)]
cent = np.zeros(shape=(5, 58,3))
for i in range(5):
cent[i] = ndimage.measurements.center_of_mass(np.array(groupby[i]))
return np.array(cent)
import mdtools.deshaw as DE
import itertools as it
import pickle, string
from collections import defaultdict, OrderedDict
import datatools.lattice as lat
import numpy as np
import os
import logging
home = os.getenv('HOME')
k_domain = string.ascii_lowercase + string.ascii_uppercase
tok = lambda x: ''.join(sorted([k_domain[i] for i in x]))
DS = 10*np.load('../data/de_ds_mu.npy')
label = DE.loadlabels_aslist()
scale = 5
blabel = [label[int(i/22.09)] for i in range(91116)]
bL = [blabel[i] for i in range(0, 90000, scale)]
D = DS[range(0, 90000, scale)]
cutoff = 8.
CM = (D<cutoff).astype(int)
K2 = lat.reduced_feature_set(CM, .02); len(K2)
F2 = [i for i in K2 if D[:,i].clip(0, cutoff).std() > .2]; len(F2)
F2std = [x[0] for x in sorted([(i, D[:,i].clip(0, cutoff).std()) for i in F2], key = lambda x: x[1])]
prs = list(it.combinations(range(58), 2))
plist = [prs[i] for i in F2std]
restrack = [[] for i in range(58)]
for a, b in plist:
restrack[a].append((a,b))
import itertools as it
from collections import defaultdict, OrderedDict
import datatools.lattice as lat
import datatools.datareduce as DR
import mdtools.timescape as TS
TBIN10 = ['%s%d'%(a,b) for a in ['W', 'T'] for b in range(5)]
# ['W0', 'W1', 'W2', 'W3', 'W4', 'T0', 'T1', 'T2', 'T3', 'T4']
REF =
afilt,hfilt = DE.FILTER['alpha'], DE.FILTER['heavy']
cent = np.load('../data/bpti-alpha-dist-centroid.npy')
topo = md.load(home+'/work/jc/serial2/de0_0/de0_0.pdb')
topoa = topo.atom_slice(topo.top.select_atom_indices('alpha'))
deL = DE.loadlabels_aslist()
bL = [deL[int(i/22.09)] for i in range(91116)]
traj, rmsd = [], []
for tnum in range(42):
p, d = DE.getHistoricalTrajectory_prot(tnum)
traj.append(md.load(d, top=p).atom_slice(afilt))
resrms = []
ccent = np.load('../data/bpti-alpha-cartesian-centroid.npy')
for tnum, tr in enumerate(detraj):
minlist = TS.TimeScape.read_log(home+'/work/timescape/desh_%02d_minima.log'%tnum)
minima = tr.slice(minlist)
minima.superpose(topoa)
for m in minima.xyz:
resrms.append(LA.norm(ref - m, axis=1))
def generateDEShawJC(fileno, frame, jcid=None):
"""Creates input parameters for a running explcit solvenet simulation
from D.E.Shaw 1ms source trajectory. Index is the global index # from the
entire 1.0375ms data set.
"""
config = systemsettings()
index = fileno * 1000 + frame
logging.debug(
"Generating new simulation coordinates from D.E.Shaw index #: %s",
str(index))
label = DE.loadlabels_aslist()
# Get a new uid (if needed)
jcuid = getUID() if jcid is None else jcid
# Prep file structure
jobdir = os.path.join(config.JOBDIR, jcuid)
# 1. Load starting coordinate from source traj and save to pdb file
filename = DE.getDEShawfilename(fileno) % fileno
src_file = os.path.join(config.workdir, 'bpti', filename)
tmpdir = 'tmp_%d%d' % (fileno, frame) #gettempdir()
logging.info('Loading Source Coordinate from file: %s (frame # %d)',
src_file, frame)
logging.info("LABELED_STATE: %d (seq # %d/4125)", label[fileno],
fileno)
coord = md.load_frame(src_file, frame, top=DE.PDB_ALL)
logging.info('Coord Loaded: %s', str(coord))
os.makedirs(tmpdir, exist_ok=True)
tmpfile = tmpdir + '/coord.pdb'
# tmpfile = 'coord.pdb'
coord.save_pdb(tmpfile)
# 2. Convert topology PDB file (in place)
logging.info(
"Converting Topology to Charmm compatiable Force Fields on: %s",
tmpfile)
DE.convert_topology(tmpfile, split_dir=tmpdir)
# 3. Prepare new job metadata
coordFile = newPdbFile = os.path.join(jobdir, '%s.pdb' % jcuid)
newPsfFile = os.path.join(jobdir, '%s.psf' % jcuid)
newsimJob = dict(
workdir=jobdir,
name=jcuid,
tmploc=tmpdir,
coord=coordFile,
pdb=newPdbFile,
psf=newPsfFile,
src_traj=src_file,
)
# Add in the force field data from global config and pre-set constraint file:
newsimJob['ffield_dir'] = os.path.join(config.workdir,
config.sim_params['ffield_dir'])
if not os.path.exists(jobdir):
os.makedirs(jobdir)
# 4. Run PSFGEN to set up start coords and structure file
logging.info(" Running PSFGen to set up simulation pdf/pdb files.")
stdout = executecmd(DE.psfgen(newsimJob))
logging.debug(" PSFGen COMPLETE!!")
# logging.debug(stdout)
# 5. Update PDB file with unitcell data and temperature control factors
logging.info('Updating Unit Cell data and setting contraints in PDB')
DE.reset_pdb(newPdbFile)
# 5. Clean up
del newsimJob['tmploc']
# shutil.rmtree(tmpdir)
return jcuid, newsimJob
def generateDEShawJC(fileno, frame, jcid=None):
"""Creates input parameters for a running explcit solvenet simulation
from D.E.Shaw 1ms source trajectory. Index is the global index # from the
entire 1.0375ms data set.
"""
config = systemsettings()
index = fileno * 1000 + frame
logging.debug("Generating new simulation coordinates from D.E.Shaw index #: %s", str(index))
label = DE.loadlabels_aslist()
# Get a new uid (if needed)
jcuid = getUID() if jcid is None else jcid
# Prep file structure
jobdir = os.path.join(config.JOBDIR, jcuid)
# 1. Load starting coordinate from source traj and save to pdb file
filename = DE.getDEShawfilename(fileno) % fileno
src_file = os.path.join(config.workdir, 'bpti', filename)
tmpdir = 'tmp_%d%d' % (fileno, frame) #gettempdir()
logging.info('Loading Source Coordinate from file: %s (frame # %d)', src_file, frame)
logging.info("LABELED_STATE: %d (seq # %d/4125)", label[fileno], fileno)
coord = md.load_frame(src_file, frame, top=DE.PDB_ALL)
logging.info('Coord Loaded: %s', str(coord))
os.makedirs(tmpdir, exist_ok=True)
tmpfile = tmpdir + '/coord.pdb'
# tmpfile = 'coord.pdb'
coord.save_pdb(tmpfile)
# 2. Convert topology PDB file (in place)
logging.info("Converting Topology to Charmm compatiable Force Fields on: %s", tmpfile)
DE.convert_topology(tmpfile, split_dir=tmpdir)
# 3. Prepare new job metadata
coordFile = newPdbFile = os.path.join(jobdir, '%s.pdb' % jcuid)
newPsfFile = os.path.join(jobdir, '%s.psf' % jcuid)
newsimJob = dict(workdir=jobdir,
name = jcuid,
tmploc = tmpdir,
coord = coordFile,
pdb = newPdbFile,
psf = newPsfFile,
src_traj = src_file,
)
# Add in the force field data from global config and pre-set constraint file:
newsimJob['ffield_dir'] = os.path.join(config.workdir, config.sim_params['ffield_dir'])
if not os.path.exists(jobdir):
os.makedirs(jobdir)
# 4. Run PSFGEN to set up start coords and structure file
logging.info(" Running PSFGen to set up simulation pdf/pdb files.")
stdout = executecmd(DE.psfgen(newsimJob))
logging.debug(" PSFGen COMPLETE!!")
# logging.debug(stdout)
# 5. Update PDB file with unitcell data and temperature control factors
logging.info('Updating Unit Cell data and setting contraints in PDB')
DE.reset_pdb(newPdbFile)
# 5. Clean up
del newsimJob['tmploc']
# shutil.rmtree(tmpdir)
return jcuid, newsimJob