def addConservation( self, pfamEntries=None, verbose=0, log=None):
"""
Adds a conservation score profile from pFam HMMs. See L{Biskit.Hmmer}
The theoretically most useful one is 'cons_ent' which gives the relative
entropy of the residue distribution with respect to the background
distribution of amino acids (Kullback-Leibler distance) in swissprot.
See PMID 16916457.
@param pfamEntries: External hmmSearch result, list of
(non-overlapping) profile hits.
(default: None, do the search) Example::
[{'ribonuclease': [[1, 108]]},..]
[{profileName : [ [startPos, endPos],
[start2, end2]]}]
- startPos, endPos as reported by hmmPfam
for PDB sequence generated from this model
@type pfamEntries: [{dict}]
@param verbose: verbosity level (default: 0)
@type verbose: 1|0
@param log: Log file for messages [STDOUT]
@type log: Biskit.LogFile
@raise ExeConfigError: if external application is missing
"""
## mask out solvent and other troublemakers
mask = self.m.maskProtein()
resmask = self.m.atom2resMask( mask )
m = self.m
if not N.alltrue( mask ):
m = self.m.compress( mask )
h = Hmmer( verbose=verbose, log=log )
h.checkHmmdbIndex()
p, hmmHits = h.scoreAbsSum( m, hmmNames=pfamEntries )
self.m.residues.set( 'cons_abs', p, hmmHits=hmmHits, mask=resmask,
comment="absolute sum of all 20 hmm scores per position",
version= T.dateString() + ' ' + self.version() )
p, hmmHits = h.scoreMaxAll( m, hmmNames=hmmHits )
self.m.residues.set( 'cons_max', p, hmmHits=hmmHits, mask=resmask,
comment="max of 20 hmm scores (-average / SD) per position",
version= T.dateString() + ' ' + self.version() )
p, hmmHits = h.scoreEntropy( m, hmmNames=hmmHits )
self.m.residues.set( 'cons_ent', p, hmmHits=hmmHits, mask=resmask,
comment="relative entropy (Kullback-Leibler distance) between "\
+"observed and background amino acid distribution "\
+"(high -> high conservation/discrimination)",
version= T.dateString() + ' ' + self.version() )
def addASA(self):
"""
Add profiles of Accessible Surface Area: 'relASA', 'ASA_total',
'ASA_sc', 'ASA_bb'. See L{Biskit.WhatIf}
@note: Using WhatIf to calculate relative accessabilities is not
nessesary any more. SurfaceRacer now also adds a profile
'relAS' (conataining relative solvent accessible surf) and
'relMS' (relative molecular surface).
@raise ProfileError: if WhatIf-returned atom/residue lists don't match.
Usually that means, WhatIf didn't recognize some
residue name
"""
w = WhatIf(self.m)
atomRelAcc, resASA, resMask = w.run()
## normalAtoms = N0.logical_not( N0.logical_or(self.m.maskHetatm(),
## self.m.maskSolvent() ) )
normalAtoms = self.m.maskProtein(standard=1)
normalRes = self.m.atom2resMask(normalAtoms)
self.m.atoms.set(
'relASA',
atomRelAcc, ## normalAtoms, 0,
comment='relative accessible surface area in %',
version=T.dateString() + ' ' + self.version())
self.m.residues.set('ASA_total',
resASA[:, 0],
normalRes,
0,
comment='accessible surface area in A^2',
version=T.dateString() + ' ' + self.version())
self.m.residues.set(
'ASA_sc',
resASA[:, 1],
normalRes,
0,
comment='side chain accessible surface area in A^2',
version=T.dateString() + ' ' + self.version())
self.m.residues.set('ASA_bb',
resASA[:, 2],
normalRes,
0,
comment='back bone accessible surface area in A^2',
version=T.dateString() + ' ' + self.version())
def addDensity( self, radius=6, minasa=None, profName='density' ):
"""
Count the number of heavy atoms within the given radius.
Values are only collected for atoms with |minasa| accessible surface
area.
@param minasa: relative exposed surface - 0 to 100%
@type minasa: float
@param radius: in Angstrom
@type radius: float
"""
mHeavy = self.m.maskHeavy()
xyz = N.compress( mHeavy, self.m.getXyz(), 0 )
if minasa and self.m.profile( 'relAS', 0 ) == 0:
self.addASA()
if minasa:
mSurf = self.m.profile2mask( 'relAS', minasa )
else:
mSurf = N.ones( self.m.lenAtoms() )
## loop over all surface atoms
surf_pos = N.nonzero( mSurf )
contacts = []
for i in surf_pos:
dist = N.sum(( xyz - self.m.xyz[i])**2, 1)
contacts += [ N.sum( N.less(dist, radius**2 )) -1]
self.m.atoms.set( profName, contacts, mSurf, default=-1,
comment='atom density radius %3.1fA' % radius,
version= T.dateString() + ' ' + self.version() )
def addSecondaryStructure(self):
"""
Adds a residue profile with the secondary structure as
calculated by the DSSP program.
Profile code::
B = residue in isolated beta-bridge
E = extended strand, participates in beta ladder
G = 3-helix (3/10 helix)
I = 5 helix (pi helix)
T = hydrogen bonded turn
S = bend
. = loop or irregular
@raise ExeConfigError: if external application is missing
"""
dssp = Dssp(self.m)
rmodel = dssp.run()
self.m.residues.set('secondary',
rmodel['dssp'],
comment='secondary structure from DSSP',
version=T.dateString() + ' ' + self.version(),
default='.')
self.m.residues.set('dssp_acc',
rmodel['dssp_acc'],
comment='accessible surface area from DSSP')
self.m.residues.set('dssp_phi',
rmodel['dssp_phi'],
comment='PHI angle from DSSP',
default=360.0)
self.m.residues.set('dssp_psi',
rmodel['dssp_psi'],
comment='PSI angle from DSSP',
default=360.0)
def addSecondaryStructure(self):
"""
Adds a residue profile with the secondary structure as
calculated by the DSSP program.
Profile code::
B = residue in isolated beta-bridge
E = extended strand, participates in beta ladder
G = 3-helix (3/10 helix)
I = 5 helix (pi helix)
T = hydrogen bonded turn
S = bend
. = loop or irregular
@raise ExeConfigError: if external application is missing
"""
prot_mask = self.m.maskProtein()
model = self.m.compress(prot_mask)
dssp = Dssp(model)
ss = dssp.run()
self.m.residues.set('secondary',
ss,
mask=self.m.atom2resMask(prot_mask),
comment='secondary structure from DSSP',
version=T.dateString() + ' ' + self.version())
def __init__(self, lst=[] ):
"""
@param lst: list of Complexes
@type lst: [Complex]
"""
## non-redundant rec/lig_models of all complexes indexed by file name
self.models = ComplexModelRegistry()
self.initVersion = t.dateString() + ' ' + self.version()
if lst != []:
self.extend( lst )
def __init__(self, lst=[] ):
"""
@param lst: list of Complexes
@type lst: [Complex]
"""
## non-redundant rec/lig_models of all complexes indexed by file name
self.models = ComplexModelRegistry()
self.initVersion = t.dateString() + ' ' + self.version()
if lst != []:
self.extend( lst )
def addASA( self ):
"""
Add profiles of Accessible Surface Area: 'relASA', 'ASA_total',
'ASA_sc', 'ASA_bb'. See L{Biskit.WhatIf}
@note: Using WhatIf to calculate relative accessabilities is not
nessesary any more. SurfaceRacer now also adds a profile
'relAS' (conataining relative solvent accessible surf) and
'relMS' (relative molecular surface).
@raise ProfileError: if WhatIf-returned atom/residue lists don't match.
Usually that means, WhatIf didn't recognize some
residue name
"""
w = WhatIf( self.m )
atomRelAcc, resASA, resMask = w.run()
## normalAtoms = N.logical_not( N.logical_or(self.m.maskHetatm(),
## self.m.maskSolvent() ) )
normalAtoms = self.m.maskProtein( standard=1 )
normalRes = self.m.atom2resMask( normalAtoms )
self.m.atoms.set( 'relASA', atomRelAcc, ## normalAtoms, 0,
comment='relative accessible surface area in %',
version= T.dateString() + ' ' + self.version() )
self.m.residues.set( 'ASA_total', resASA[:,0], normalRes, 0,
comment='accessible surface area in A^2',
version= T.dateString() + ' ' + self.version() )
self.m.residues.set( 'ASA_sc', resASA[:,1], normalRes, 0,
comment='side chain accessible surface area in A^2',
version= T.dateString() + ' ' + self.version() )
self.m.residues.set( 'ASA_bb', resASA[:,2], normalRes, 0,
comment='back bone accessible surface area in A^2',
version= T.dateString() + ' ' + self.version() )
def addSurfaceMask( self, pname='relAS' ):
"""
Adds a surface mask profie that contains atoms with > 40% exposure
compared to a random coil state.
@param pname: name of relative profile to use
(Whatif-relASA OR SurfaceRacer - relAS)
(default: relAS)
@type pname: str
"""
r = self.m.profile2mask( pname, cutoff_min=40 )
self.m.residues.set( 'surfMask', self.m.atom2resMask(r),
comment='residues with any atom > 40% exposed',
version= T.dateString() + ' ' + self.version() )
def addSurfaceMask(self, pname='relAS'):
"""
Adds a surface mask profie that contains atoms with > 40% exposure
compared to a random coil state.
@param pname: name of relative profile to use
(Whatif-relASA OR SurfaceRacer - relAS)
(default: relAS)
@type pname: str
"""
r = self.m.profile2mask(pname, cutoff_min=40)
self.m.residues.set('surfMask',
self.m.atom2resMask(r),
comment='residues with any atom > 40% exposed',
version=T.dateString() + ' ' + self.version())
def addSecondaryStructure( self ):
"""
Adds a residue profile with the secondary structure as
calculated by the DSSP program.
Profile code::
B = residue in isolated beta-bridge
E = extended strand, participates in beta ladder
G = 3-helix (3/10 helix)
I = 5 helix (pi helix)
T = hydrogen bonded turn
S = bend
. = loop or irregular
@raise ExeConfigError: if external application is missing
"""
dssp = Dssp( self.m )
ss = dssp.run()
self.m.residues.set( 'secondary', ss,
comment='secondary structure from DSSP',
version= T.dateString() + ' ' + self.version() )
def addDensity(self, radius=6, minasa=None, profName='density'):
"""
Count the number of heavy atoms within the given radius.
Values are only collected for atoms with |minasa| accessible surface
area.
@param minasa: relative exposed surface - 0 to 100%
@type minasa: float
@param radius: in Angstrom
@type radius: float
"""
mHeavy = self.m.maskHeavy()
xyz = N0.compress(mHeavy, self.m.getXyz(), 0)
if minasa and self.m.profile('relAS', 0) == 0:
self.addASA()
if minasa:
mSurf = self.m.profile2mask('relAS', minasa)
else:
mSurf = N0.ones(self.m.lenAtoms())
## loop over all surface atoms
surf_pos = N0.nonzero(mSurf)
contacts = []
for i in surf_pos:
dist = N0.sum((xyz - self.m.xyz[i])**2, 1)
contacts += [N0.sum(N0.less(dist, radius**2)) - 1]
self.m.atoms.set(profName,
contacts,
mSurf,
default=-1,
comment='atom density radius %3.1fA' % radius,
version=T.dateString() + ' ' + self.version())
def __init__(self):
"""
Override but call.
"""
self.initVersion = t.dateString() + ' ' + self.version()
def addSurfaceRacer( self, probe=1.4, vdw_set=1, probe_suffix=0, mask=None ):
"""
Always adds three different profiles as calculated by fastSurf::
curvature - average curvature (or curvature_1.4 if probe_suffix=1)
MS - molecular surface area (or MS_1.4 if probe_suffix=1)
AS - accessible surface area (or AS_1.4 if probe_suffix=1)
If the probe radii is 1.4 Angstrom and the Richards vdw radii
set is used the following two profiles are also added::
relAS - Relative solvent accessible surface
relMS - Relative molecular surface
See {Biskit.SurfaceRacer}
@param probe: probe radius
@type probe: float
@param vdw_set: defines what wdv-set to use (1-Richards, 2-Chothia)
@type vdw_set: 1|2
@param probe_suffix: append probe radius to profile names
@type probe_suffix: 1|0
@param mask: optional atom mask to apply before calling surface racer
(default: heavy atoms AND NOT solvent)
@type mask: [ bool ]
@raise ExeConfigError: if external application is missing
"""
name_MS = 'MS' + probe_suffix * ('_%3.1f' % probe)
name_AS = 'AS' + probe_suffix * ('_%3.1f' % probe)
name_curv = 'curvature' + probe_suffix * ('_%3.1f' % probe)
## hydrogens + waters are not allowed during FastSurf calculation
mask = mask if mask is not None else \
self.m.maskHeavy() * N.logical_not( self.m.maskSolvent() )
fs = SurfaceRacer( self.m, probe, vdw_set=vdw_set, mask=mask )
fs_dic = fs.run()
fs_info= fs_dic['surfaceRacerInfo']
self.m.atoms.set( name_MS, fs_dic['MS'], mask, 0,
comment='Molecular Surface area in A',
version= T.dateString() + ' ' + self.version(),
**fs_info )
self.m.atoms.set( name_AS, fs_dic['AS'], mask, 0,
comment='Accessible Surface area in A',
version= T.dateString() + ' ' + self.version(),
**fs_info )
self.m.atoms.set( name_curv, fs_dic['curvature'], mask, 0,
comment='Average curvature',
version= T.dateString() + ' ' + self.version(),
**fs_info )
if round(probe, 1) == 1.4 and vdw_set == 1 and 'relAS' in fs_dic:
self.m.atoms.set( 'relAS', fs_dic['relAS'], mask, 0,
comment='Relative solvent accessible surf.',
version= T.dateString()+' ' +self.version(),
**fs_info )
self.m.atoms.set( 'relMS', fs_dic['relMS'], mask, 0,
comment='Relative molecular surf.',
version= T.dateString()+' '+self.version(),
**fs_info )
def addSurfaceRacer(self, probe=1.4, vdw_set=1, probe_suffix=0, mask=None):
"""
Always adds three different profiles as calculated by fastSurf::
curvature - average curvature (or curvature_1.4 if probe_suffix=1)
MS - molecular surface area (or MS_1.4 if probe_suffix=1)
AS - accessible surface area (or AS_1.4 if probe_suffix=1)
If the probe radii is 1.4 Angstrom and the Richards vdw radii
set is used the following two profiles are also added::
relAS - Relative solvent accessible surface
relMS - Relative molecular surface
See {Biskit.SurfaceRacer}
@param probe: probe radius
@type probe: float
@param vdw_set: defines what wdv-set to use (1-Richards, 2-Chothia)
@type vdw_set: 1|2
@param probe_suffix: append probe radius to profile names
@type probe_suffix: 1|0
@param mask: optional atom mask to apply before calling surface racer
(default: heavy atoms AND NOT solvent)
@type mask: [ bool ]
@raise ExeConfigError: if external application is missing
"""
name_MS = 'MS' + probe_suffix * ('_%3.1f' % probe)
name_AS = 'AS' + probe_suffix * ('_%3.1f' % probe)
name_curv = 'curvature' + probe_suffix * ('_%3.1f' % probe)
## hydrogens + waters are not allowed during FastSurf calculation
mask = mask if mask is not None else \
self.m.maskHeavy() * N0.logical_not( self.m.maskSolvent() )
fs = SurfaceRacer(self.m, probe, vdw_set=vdw_set, mask=mask)
fs_dic = fs.run()
fs_info = fs_dic['surfaceRacerInfo']
self.m.atoms.set(name_MS,
fs_dic['MS'],
mask,
0,
comment='Molecular Surface area in A',
version=T.dateString() + ' ' + self.version(),
**fs_info)
self.m.atoms.set(name_AS,
fs_dic['AS'],
mask,
0,
comment='Accessible Surface area in A',
version=T.dateString() + ' ' + self.version(),
**fs_info)
self.m.atoms.set(name_curv,
fs_dic['curvature'],
mask,
0,
comment='Average curvature',
version=T.dateString() + ' ' + self.version(),
**fs_info)
if round(probe, 1) == 1.4 and vdw_set == 1 and 'relAS' in fs_dic:
self.m.atoms.set('relAS',
fs_dic['relAS'],
mask,
0,
comment='Relative solvent accessible surf.',
version=T.dateString() + ' ' + self.version(),
**fs_info)
self.m.atoms.set('relMS',
fs_dic['relMS'],
mask,
0,
comment='Relative molecular surf.',
version=T.dateString() + ' ' + self.version(),
**fs_info)
def addConservation(self, pfamEntries=None, verbose=0, log=None):
"""
Adds a conservation score profile from pFam HMMs. See L{Biskit.Hmmer}
The theoretically most useful one is 'cons_ent' which gives the relative
entropy of the residue distribution with respect to the background
distribution of amino acids (Kullback-Leibler distance) in swissprot.
See PMID 16916457.
@param pfamEntries: External hmmSearch result, list of
(non-overlapping) profile hits.
(default: None, do the search) Example::
[{'ribonuclease': [[1, 108]]},..]
[{profileName : [ [startPos, endPos],
[start2, end2]]}]
- startPos, endPos as reported by hmmPfam
for PDB sequence generated from this model
@type pfamEntries: [{dict}]
@param verbose: verbosity level (default: 0)
@type verbose: 1|0
@param log: Log file for messages [STDOUT]
@type log: Biskit.LogFile
@raise ExeConfigError: if external application is missing
"""
## mask out solvent and other troublemakers
mask = self.m.maskProtein()
resmask = self.m.atom2resMask(mask)
m = self.m
if not N0.alltrue(mask):
m = self.m.compress(mask)
h = Hmmer(verbose=verbose, log=log)
h.checkHmmdbIndex()
p, hmmHits = h.scoreAbsSum(m, hmmNames=pfamEntries)
self.m.residues.set(
'cons_abs',
p,
hmmHits=hmmHits,
mask=resmask,
comment="absolute sum of all 20 hmm scores per position",
version=T.dateString() + ' ' + self.version())
p, hmmHits = h.scoreMaxAll(m, hmmNames=hmmHits)
self.m.residues.set(
'cons_max',
p,
hmmHits=hmmHits,
mask=resmask,
comment="max of 20 hmm scores (-average / SD) per position",
version=T.dateString() + ' ' + self.version())
p, hmmHits = h.scoreEntropy(m, hmmNames=hmmHits)
self.m.residues.set( 'cons_ent', p, hmmHits=hmmHits, mask=resmask,
comment="relative entropy (Kullback-Leibler distance) between "\
+"observed and background amino acid distribution "\
+"(high -> high conservation/discrimination)",
version= T.dateString() + ' ' + self.version() )
