def motif_from_res(self, res, bps):
m = motif.Motif()
chains = chain.connect_residues_into_chains(res)
m.structure.chains = chains
m.basepairs = bps
ends = self._setup_basepair_ends(m.structure, bps)
m.ends = ends
self._setup_secondary_structure(m)
return m
def motif_from_chains(self, chains, bps):
m = motif.Motif()
m.structure.chains = chains
m.basepairs = bps
ends = self._setup_basepair_ends(m.structure, bps)
m.ends = ends
try:
self._setup_secondary_structure(m)
except:
pass
return m
def motif_from_bps(self, bps):
m = motif.Motif()
res = []
for bp in bps:
res.extend(bp.residues())
chains = chain.connect_residues_into_chains(res)
m.structure.chains = chains
m.basepairs = bps
m.ends = self._setup_basepair_ends(m.structure, m.basepairs)
self._setup_secondary_structure(m)
return m
def motif_from_file(self, path, include_protein=0):
filename = util.filename(path)
# is a motif directory
if os.path.isdir(path):
s = structure.structure_from_pdb(path + "/" + filename + ".pdb")
else:
s = structure.structure_from_pdb(path)
filename = filename[:-4]
basepairs = rna_structure.basepairs_from_x3dna(path, filename, s)
ends = rna_structure.ends_from_basepairs(s, basepairs)
r_struct = rna_structure.RNAStructure(s, basepairs, ends, filename,
path, motif_type.UNKNOWN)
#if len(r_struct.residues()) == 0:
# raise exceptions.MotifFactoryException(
# )
m = motif.Motif(r_struct)
m.score = self.scorer.score(m)
#try:
self._setup_secondary_structure(m)
#except:
# print "did not parse secondary_structure", m.name
if include_protein:
if os.path.isdir(path):
p_residues = pdb_parser.parse(path + "/" + filename + ".pdb",
protein=1, rna=0)
else:
p_residues = pdb_parser.parse(path, protein=1, rna=0)
beads = []
for r in p_residues:
a = r.get_atom("CA")
beads.append(residue.Bead(a.coords, residue.BeadType.BASE))
m.protein_beads = beads
return m
def mdmodule(self, motif_widths=(7, 10, 13, 16, 19), width=600):
"""Run a de novo motif scan of ChIP regions.
De novo motifs will be returned as a MotifList object.
Adapted from Cliff Meyer's ChIP_region.MDscan() method.
Directly calls Xiaole Shirley Liu's _MDmod program."""
motifs = motif.MotifList()
# prepare input sequences
input = self.copy()
input.unique()
input.trim(int(width / 2))
input.read_sequence(True)
#OBSOLETE: DELETE!!
# prepare background sequences
background = self.copy()
background.unique()
background.trim(int(width))
background.read_sequence(True)
# scan for motifs
for motif_width in motif_widths:
raw_pssms = MDmod(i=input.sequence,
b=background.sequence,
w=motif_width,
t=50,
s=50,
n=100,
r=10)
for raw_pssm in raw_pssms:
motif_id = 'denovo%d' % len(motifs)
tmp = motif.Motif()
tmp.id = motif_id
tmp.pssm = raw_pssm
motifs.append(tmp)
return motifs
def _ConvertToOldMotif(self, motifid):
#import mdseqpos.motif as motif
p = motif.Motif()
p = p.from_dict(self.motifs[motifid])
return p
def pose_from_motif_tree_new(self, structure, basepairs, nodes,
designable):
p = pose.Pose()
p.designable = designable
p.name = "assembled"
p.path = "assembled"
p.structure = structure
p.basepairs = basepairs
p.ends = motif_factory.factory._setup_basepair_ends(
structure, basepairs)
ss = secondary_structure_factory.factory.secondary_structure_from_motif(
p)
p.secondary_structure = ss
p.end_ids = ["" for x in p.ends]
for i, end in enumerate(p.ends):
res1 = ss.get_residue(uuid=end.res1.uuid)
res2 = ss.get_residue(uuid=end.res2.uuid)
ss_end = ss.get_bp(res1, res2)
p.end_ids[i] = secondary_structure.assign_end_id(ss, ss_end)
#update all motifs in pose, make sure that the correct basepairs are used
#since basepairs are used to align there are two choices of which basepair to
#be used
p_res = p.residues()
for j, n in enumerate(nodes):
m = n.data
residue_map = {}
print j
chains = []
res = []
for c in m.chains():
bounds = [0, len(c.residues)]
first = 0
start_pos = 0
found = 0
for i, r in enumerate(c.residues):
r_new = p.get_residue(uuid=r.uuid)
if r_new is not None:
first = i
start_pos = p_res.index(r_new)
found = 1
break
bounds[0] = start_pos - first
bounds[1] = bounds[0] + len(c.residues)
print bounds[0], bounds[1]
chains.append(chain.Chain(p_res[bounds[0]:bounds[1]]))
res.extend(chains[-1].residues)
bps = self._subselect_bps_from_res(p.basepairs, res)
m_copy = motif_factory.factory.motif_from_chains(chains, bps)
m_copy.to_pdb("m." + str(j) + ".pdb")
m.to_pdb("org." + str(j) + ".pdb")
continue
exit()
if len(bps) != len(m.basepairs):
raise ValueError(
"something went horribly wrong: did not find all basepairs"
)
m_copy = motif.Motif()
m_copy.mtype = m.mtype
m_copy.name = m.name
if m_copy.mtype is not motif_type.HELIX:
best_ends = []
best_end_ids = []
for i, end in enumerate(m.ends):
best = 1000
best_end = None
best_end_id = None
for c_end in m_copy.ends:
dist = util.distance(end.d(), c_end.d())
if dist < best:
best_end = c_end
best_end_id = m.end_ids[i]
best = dist
best_ends.append(best_end)
best_end_ids.append(best_end_id)
m_copy.ends = best_ends
m_copy.end_ids = best_end_ids
p.motif_dict[m_copy.mtype].append(m_copy)
p.motif_dict[motif_type.ALL].append(m_copy)
self._add_secondary_structure_motifs(p)
self.standardize_pose(p)
return p
def pose_from_motif_tree(self, structure, basepairs, nodes, designable):
p = pose.Pose()
p.designable = designable
p.name = "assembled"
p.path = "assembled"
p.structure = structure
p.basepairs = basepairs
p.ends = motif_factory.factory._setup_basepair_ends(
structure, basepairs)
ss = secondary_structure_factory.factory.secondary_structure_from_motif(
p)
p.secondary_structure = ss
p.end_ids = ["" for x in p.ends]
for i, end in enumerate(p.ends):
res1 = ss.get_residue(uuid=end.res1.uuid)
res2 = ss.get_residue(uuid=end.res2.uuid)
ss_end = ss.get_bp(res1, res2)
p.end_ids[i] = secondary_structure.assign_end_id(ss, ss_end)
#update all motifs in pose, make sure that the correct basepairs are used
#since basepairs are used to align there are two choices of which basepair to
#be used
for j, n in enumerate(nodes):
m = n.data
bps = []
residue_map = {}
for bp in m.basepairs:
new_bp = p.get_basepair(bp_uuid=bp.uuid)
if len(new_bp) != 0:
bps.append(new_bp[0])
continue
best = 1000
best_bp = None
for c in n.connections:
if c is None:
continue
if bp not in m.ends:
raise ValueError("cannot find a non end basepair")
if c.end_index(n.index) != m.ends.index(bp):
continue
partner = c.partner(n.index)
alt_bp = partner.data.ends[c.end_index(partner.index)]
pose_bps = p.get_basepair(bp_uuid=alt_bp.uuid)
if len(pose_bps) == 0:
raise ValueError("cannot find partner end bp in pose")
#TODO Check if this is okay, not sure if I have to try and match
#TODO residues correctly
residue_map[bp.res1] = pose_bps[0].res1
residue_map[bp.res2] = pose_bps[0].res2
bps.append(pose_bps[0])
break
chains = []
for c in m.chains():
res = []
for r in c.residues:
new_r = p.get_residue(uuid=r.uuid)
if new_r is not None:
res.append(new_r)
elif r in residue_map:
res.append(residue_map[r])
else:
print r, r.uuid
raise ValueError("cannot find residue")
chains.append(chain.Chain(res))
if len(bps) != len(m.basepairs):
raise ValueError(
"something went horribly wrong: did not find all basepairs"
)
m_copy = motif.Motif()
m_copy.mtype = m.mtype
m_copy.name = m.name
m_copy.structure.chains = chains
m_copy.basepairs = bps
m_copy.ends = motif_factory.factory._setup_basepair_ends(
m_copy.structure, bps)
motif_factory.factory._setup_secondary_structure(m_copy)
if m_copy.mtype is not motif_type.HELIX:
best_ends = []
best_end_ids = []
for i, end in enumerate(m.ends):
best = 1000
best_end = None
best_end_id = None
for c_end in m_copy.ends:
dist = util.distance(end.d(), c_end.d())
if dist < best:
best_end = c_end
best_end_id = m.end_ids[i]
best = dist
best_ends.append(best_end)
best_end_ids.append(best_end_id)
m_copy.ends = best_ends
m_copy.end_ids = best_end_ids
p.motif_dict[m_copy.mtype].append(m_copy)
p.motif_dict[motif_type.ALL].append(m_copy)
self._add_secondary_structure_motifs(p)
self.standardize_pose(p)
return p
def pose_from_motif_tree_old(self, structure, basepairs, motifs,
designable):
p = pose.Pose()
p.designable = designable
p.name = "assembled"
p.path = "assembled"
p.structure = structure
p.basepairs = basepairs
p.ends = motif_factory.factory._setup_basepair_ends(
structure, basepairs)
ss = secondary_structure_factory.factory.secondary_structure_from_motif(
p)
p.secondary_structure = ss
p.end_ids = ["" for x in p.ends]
for i, end in enumerate(p.ends):
res1 = ss.get_residue(uuid=end.res1.uuid)
res2 = ss.get_residue(uuid=end.res2.uuid)
ss_end = ss.get_bp(res1, res2)
p.end_ids[i] = secondary_structure.assign_end_id(ss, ss_end)
#update all motifs in pose, make sure that the correct basepairs are used
#since basepairs are used to align there are two choices of which basepair to
#be used
for j, m in enumerate(motifs):
bps = []
residue_map = {}
for bp in m.basepairs:
new_bp = p.get_basepair(bp_uuid=bp.uuid)
#print bp.residues(), len(new_bp)
if len(new_bp) != 0:
bps.append(new_bp[0])
continue
best = 1000
best_bp = None
for m2 in motifs:
if m == m2:
continue
for end in m2.ends:
alt_bp = p.get_basepair(bp_uuid=end.uuid)
if len(alt_bp) == 0:
continue
dist = util.distance(bp.d(), alt_bp[0].d())
if dist < best:
best_bp = alt_bp[0]
best = dist
if best_bp is None:
continue
new_bp = [best_bp]
for r1 in best_bp.residues():
r1_cent = util.center(r1.atoms)
best_match = None
best_dist = 1000
for r2 in bp.residues():
r2_cent = util.center(r2.atoms)
dist = util.distance(r1_cent, r2_cent)
if dist < best_dist:
best_dist = dist
best_match = r2
residue_map[best_match] = r1
bps.append(new_bp[0])
chains = []
#print m.name
for c in m.chains():
res = []
for r in c.residues:
new_r = p.get_residue(uuid=r.uuid)
if new_r is not None:
res.append(new_r)
elif r in residue_map:
res.append(residue_map[r])
else:
print r, r.uuid
raise ValueError("cannot find residue")
chains.append(chain.Chain(res))
if len(bps) != len(m.basepairs):
raise ValueError(
"something went horribly wrong: did not find all basepairs"
)
m_copy = motif.Motif()
m_copy.mtype = m.mtype
m_copy.name = m.name
m_copy.structure.chains = chains
m_copy.basepairs = bps
m_copy.ends = motif_factory.factory._setup_basepair_ends(
m_copy.structure, bps)
motif_factory.factory._setup_secondary_structure(m_copy)
if m_copy.mtype is not motif_type.HELIX:
best_ends = []
best_end_ids = []
for i, end in enumerate(m.ends):
best = 1000
best_end = None
best_end_id = None
for c_end in m_copy.ends:
dist = util.distance(end.d(), c_end.d())
if dist < best:
best_end = c_end
best_end_id = m.end_ids[i]
best = dist
best_ends.append(best_end)
best_end_ids.append(best_end_id)
m_copy.ends = best_ends
m_copy.end_ids = best_end_ids
p.motif_dict[m_copy.mtype].append(m_copy)
p.motif_dict[motif_type.ALL].append(m_copy)
self._add_secondary_structure_motifs(p)
self.standardize_pose(p)
return p