def from_prody_atomgroup(
cls,
name: ProteinKey,
protein: pd.AtomGroup,
split_type: SplitType = SplitType.KMER,
split_size: int = 16,
selection: str = "calpha",
upsample_rate: int = 50,
moment_types: List[MomentType] = (
MomentType.O_3,
MomentType.O_4,
MomentType.O_5,
MomentType.F,
),
):
"""
Construct MomentInvariants instance from a ProDy AtomGroup object.
Selects according to `selection` string, (default = alpha carbons)
`moment_types` determines which moments are calculated.
Example
--------
>>> invariants = MomentInvariants.from_prody_atomgroup(atom_group, split_type=SplitType.RADIUS, moment_types=[MomentType.O_3, MomentType.F, MomentType.phi_7, MomentType.phi_12])
"""
protein: pd.AtomGroup = protein.select("protein").select(selection)
coordinates: np.ndarray = protein.getCoords()
residue_splits = group_indices(protein.getResindices())
shape = cls(
name,
len(residue_splits),
coordinates,
residue_splits,
protein.getIndices(),
sequence=protein.getSequence(),
split_type=split_type,
split_size=split_size,
upsample_rate=upsample_rate,
moment_types=moment_types,
)
shape._split(split_type)
return shape
def visualizemapApp(inp_file, out_file, sel_type, atom_group: prody.AtomGroup,
vmin_fltr, vmax_fltr, dmin_fltr, dmax_fltr, cyl_rad):
##########################################################################
selectedAtoms = atom_group.select('protein and name CA')
##########################################################################
minColorBarLimit = 0.0
maxColorBarLimit = 1.0
# Read data file and assign to a numpy array
if sel_type.lower() == "ndcc":
# Check if the data type is sparse matrix
data_file = open(inp_file, 'r')
allLines = data_file.readlines()
data_file.close()
# Read the first line to determine if the matrix is sparse format
words = allLines[0].split()
# Read the 1st line and check if it has three columns
if (len(words) == 3):
ccMatrix = parseSparseCorrData(inp_file, selectedAtoms, \
Ctype=True,
symmetric=True,
writeAllOutput=False)
else:
ccMatrix = np.loadtxt(inp_file, dtype=float)
# Check the data range in the matrix.
minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
if minCorrelationValue < 0.0:
# Assume that it is an nDCC file
minColorBarLimit = -1.0
if maxCorrelationValue > 1.00001:
print("This correlation map is not normalized!")
# TODO: At this point, one can ask the user if s/he wants to normalize it!
sys.exit(-1)
else:
maxColorBarLimit = 1.0
elif sel_type.lower() == "dcc":
# Check if the data type is sparse matrix
data_file = open(inp_file, 'r')
allLines = data_file.readlines()
data_file.close()
# Read the first line to determine if the matrix is sparse format
words = allLines[0].split()
# Read the 1st line and check if it has three columns
if (len(words) == 3):
ccMatrix = parseSparseCorrData(inp_file, selectedAtoms, \
Ctype=True,
symmetric=True,
writeAllOutput=False)
else:
ccMatrix = np.loadtxt(inp_file, dtype=float)
# Check the data range in the matrix.
minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
minColorBarLimit = minCorrelationValue
maxColorBarLimit = maxCorrelationValue
elif sel_type.lower() == "absndcc":
# Check if the data type is sparse matrix
data_file = open(inp_file, 'r')
allLines = data_file.readlines()
data_file.close()
# Read the first line to determine if the matrix is sparse format
words = allLines[0].split()
# Read the 1st line and check if it has three columns
if (len(words) == 3):
ccMatrix = np.absolute(parseSparseCorrData(inp_file, selectedAtoms, \
Ctype=True,
symmetric=True,
writeAllOutput=False))
else:
ccMatrix = np.absolute(np.loadtxt(inp_file, dtype=float))
minColorBarLimit = 0.0
maxColorBarLimit = 1.0
elif sel_type.lower() == "lmi":
# Check if the data type is sparse matrix
data_file = open(inp_file, 'r')
allLines = data_file.readlines()
data_file.close()
# Read the first line to determine if the matrix is sparse format
words = allLines[0].split()
# Read the 1st line and check if it has three columns
if (len(words) == 3):
ccMatrix = parseSparseCorrData(inp_file, selectedAtoms, \
Ctype=True,
symmetric=True,
writeAllOutput=False)
else:
ccMatrix = convertLMIdata2Matrix(inp_file, writeAllOutput=False)
minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
minColorBarLimit = minCorrelationValue
maxColorBarLimit = maxCorrelationValue
#minColorBarLimit = 0.0
elif sel_type.lower() == "nlmi":
# Check if the data type is sparse matrix
data_file = open(inp_file, 'r')
allLines = data_file.readlines()
data_file.close()
# Read the first line to determine if the matrix is sparse format
words = allLines[0].split()
# Read the 1st line and check if it has three columns
if (len(words) == 3):
ccMatrix = parseSparseCorrData(inp_file, selectedAtoms, \
Ctype=True,
symmetric=True,
writeAllOutput=False)
else:
ccMatrix = convertLMIdata2Matrix(inp_file, writeAllOutput=False)
#minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
minColorBarLimit = 0.0
# Ideally, it is supposed to be 1 but I used 1.00001 to avoid
# rounding problems
if maxCorrelationValue > 1.00001:
print("This LMI map is not normalized!")
# TODO: At this point, one can ask the user if s/he wants to normalize it!
sys.exit(-1)
else:
maxColorBarLimit = 1.0
elif sel_type.lower() == "coeviz":
ccMatrix = np.loadtxt(inp_file, dtype=float)
minColorBarLimit = 0.0
maxColorBarLimit = 1.0
elif sel_type.lower() == "evcouplings":
ccMatrix = parseEVcouplingsScores(inp_file, selectedAtoms, False)
minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
minColorBarLimit = minCorrelationValue
maxColorBarLimit = maxCorrelationValue
elif sel_type.lower() == "generic":
# Check if the data type is sparse matrix
data_file = open(inp_file, 'r')
allLines = data_file.readlines()
data_file.close()
# Read the first line to determine if the matrix is sparse format
words = allLines[0].split()
# Read the 1st line and check if it has three columns
if (len(words) == 3):
ccMatrix = parseSparseCorrData(inp_file, selectedAtoms, \
Ctype=True,
symmetric=True,
writeAllOutput=False)
else:
ccMatrix = np.loadtxt(inp_file, dtype=float)
minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
minColorBarLimit = minCorrelationValue
maxColorBarLimit = maxCorrelationValue
elif sel_type.lower() == "eg":
# The data type is elasticity graph
ccMatrix = parseElasticityGraph(inp_file, selectedAtoms, \
writeAllOutput=False)
minCorrelationValue = np.min(ccMatrix)
maxCorrelationValue = np.max(ccMatrix)
minColorBarLimit = minCorrelationValue
maxColorBarLimit = maxCorrelationValue
else:
print(
"@> ERROR: Unknown data type: Type can only be ndcc, absndcc, lmi,\n"
)
print(
"@> coeviz or evcouplings. If you have your data in full \n"
)
print(
"@> matrix format and your data type is none of the options\n"
)
print("@> mentionned, you can set data type 'generic'.\n")
sys.exit(-1)
# Set vmin_fltr and vmax_fltr
if (vmin_fltr == None):
vmin_fltr = minColorBarLimit
if (vmax_fltr == None):
vmax_fltr = maxColorBarLimit
print(f"""@> Min. value filter: {vmin_fltr}""")
print(f"""@> Max. value filter: {vmax_fltr}""")
##########################################################################
# Call overall correlation calculation
overallCorrelationMap(ccMatrix, minColorBarLimit, maxColorBarLimit,
out_file, " ", selectedAtoms)
plotDistributions = True
VMDcylinderRadiusScale = 0.5
PMLcylinderRadiusScale = 0.3
if (cyl_rad == None):
if sel_type.lower() == "evcouplings":
VMDcylinderRadiusScale = 0.02
PMLcylinderRadiusScale = 0.02
else:
VMDcylinderRadiusScale = 0.5
PMLcylinderRadiusScale = 0.3
print(f"""@> VMD Cylinder radius: {VMDcylinderRadiusScale}""")
print(f"""@> PyMol Cylinder radius: {PMLcylinderRadiusScale}""")
else:
VMDcylinderRadiusScale = float(cyl_rad)
PMLcylinderRadiusScale = float(cyl_rad)
print(f"""@> Cylinder radius: {cyl_rad}""")
if plotDistributions:
if sel_type.lower() == "ndcc":
distanceDistribution(ccMatrix,
out_file,
"nDCC",
selectedAtoms,
absoluteValues=False,
writeAllOutput=True)
elif sel_type.lower() == "dcc":
distanceDistribution(ccMatrix,
out_file,
"DCC",
selectedAtoms,
absoluteValues=False,
writeAllOutput=True)
elif sel_type.lower() == "absndcc":
distanceDistribution(ccMatrix,
out_file,
"Abs(nDCC)",
selectedAtoms,
absoluteValues=True,
writeAllOutput=False)
elif sel_type.lower() == "lmi":
distanceDistribution(ccMatrix,
out_file,
"LMI",
selectedAtoms,
absoluteValues=True,
writeAllOutput=True)
elif sel_type.lower() == "nlmi":
distanceDistribution(ccMatrix,
out_file,
"nLMI",
selectedAtoms,
absoluteValues=True,
writeAllOutput=True)
elif sel_type.lower() == "coeviz":
distanceDistribution(ccMatrix,
out_file,
"CoeViz",
selectedAtoms,
absoluteValues=True,
writeAllOutput=True)
elif sel_type.lower() == "evcouplings":
distanceDistribution(ccMatrix,
out_file,
"EVcoupling Score",
selectedAtoms,
absoluteValues=False,
writeAllOutput=True)
elif sel_type.lower() == "generic":
distanceDistribution(ccMatrix,
out_file,
"Correlation",
selectedAtoms,
absoluteValues=False,
writeAllOutput=True)
elif sel_type.lower() == "eg":
distanceDistribution(ccMatrix,
out_file,
"Force Constants",
selectedAtoms,
absoluteValues=False,
writeAllOutput=True)
else:
print("Warning: Unknows correlation data.\n")
print(" Correlations can be dcc, ndcc, absndcc, lmi,\n")
print(" nlmi, coeviz or evcouplings!\n")
##########################################################################
# Check number of chains. If there are multiple chains, plot inter and
# intra chain correlations
chains = Counter(selectedAtoms.getChids()).keys()
saveMatrix = False
plotChains = True
if len(chains) > 1 and plotChains:
intraChainCorrelationMaps(ccMatrix, minColorBarLimit, maxColorBarLimit,
out_file, " ", selectedAtoms, saveMatrix)
interChainCorrelationMaps(ccMatrix, minColorBarLimit, maxColorBarLimit,
out_file, " ", selectedAtoms, saveMatrix)
# Here, we can filter some correlation values closer than a distance.
# Typically, it is supposed to filter out the correlation within the
# same secondary structure etc.
filterByDistance = True
if filterByDistance:
disMinValue = float(dmin_fltr)
disMaxValue = float(dmax_fltr)
ccMatrix = filterCorrelationMapByDistance(ccMatrix,
out_file,
" ",
selectedAtoms,
disMinValue,
disMaxValue,
absoluteValues=False,
writeAllOutput=False)
# Overall projection
projectCorrelationsOntoProteinVMD(
out_file,
ccMatrix,
out_file,
selectedAtoms,
vminFilter=float(vmin_fltr),
vmaxFilter=float(vmax_fltr),
cylinderRadiusScaler=VMDcylinderRadiusScale,
absoluteValues=True,
writeAllOutput=True)
projectCorrelationsOntoProteinPyMol(
out_file,
ccMatrix,
out_file,
selectedAtoms,
vminFilter=float(vmin_fltr),
vmaxFilter=float(vmax_fltr),
cylinderRadiusScaler=PMLcylinderRadiusScale,
absoluteValues=True,
writeAllOutput=True)