def mutate_residue(pose, mutant_position, mutant_aa, pack_radius,
pack_scorefxn):
if pose.is_fullatom() == False:
IOError('mutate_residue only works with fullatom poses')
test_pose = Pose()
test_pose.assign(pose)
# Create a packer task (standard)
task = TaskFactory.create_packer_task(test_pose)
# the Vector1 of booleans (a specific object) is needed for specifying the
# mutation, this demonstrates another more direct method of setting
# PackerTask options for design
aa_bool = vector1_bool()
# PyRosetta uses several ways of tracking amino acids (ResidueTypes)
# the numbers 1-20 correspond individually to the 20 proteogenic amino acids
# aa_from_oneletter returns the integer representation of an amino acid
# from its one letter code
# convert mutant_aa to its integer representation
mutant_aa = aa_from_oneletter_code(mutant_aa)
# mutation is performed by using a PackerTask with only the mutant
# amino acid available during design
# to do this, construct a Vector1 of booleans indicating which amino acid
# (by its numerical designation, see above) to allow
for i in range(1, 21):
# in Python, logical expression are evaluated with priority, thus the
# line below appends to aa_bool the truth (True or False) of the
# statement i == mutant_aa
aa_bool.append(i == mutant_aa)
# modify the mutating residue's assignment in the PackerTask using the
# Vector1 of booleans across the proteogenic amino acids
task.nonconst_residue_task(mutant_position).restrict_absent_canonical_aas(
aa_bool)
# prevent residues from packing by setting the per-residue "options" of
# the PackerTask
center = pose.residue(mutant_position).nbr_atom_xyz()
for i in range(1, pose.total_residue() + 1):
dist = center.distance_squared(test_pose.residue(i).nbr_atom_xyz())
# only pack the mutating residue and any within the pack_radius
if i != mutant_position and dist > pow(float(pack_radius), 2):
task.nonconst_residue_task(i).prevent_repacking()
# apply the mutation and pack nearby residues
packer = PackRotamersMover(pack_scorefxn, task)
packer.apply(test_pose)
return test_pose
def load_pose( pose_filename ):
'''
Load pose from a filename
:param pose_filename: str( /path/to/pose/filename )
:return: a Rosetta Pose
'''
# imports
from rosetta import Pose, pose_from_file, FoldTree
# create Pose object from filename
pose = Pose()
pose_from_file( pose, pose_filename )
# clean up the name of the pose
pose_name = pose.pdb_info().name()
pose_name = pose_name.split( '/' )[-1]
pose.pdb_info().name( pose_name )
# store the original FoldTree and add empty loops for use later
pose.orig_fold_tree = FoldTree( pose.fold_tree() )
pose.loops = None
pose.loops_file = None
return pose
def Fc_glycan_rmsd( working, working_Fc_glycan_chains, native, native_Fc_glycan_chains, decoy_num, dump_dir ):
"""
:param working: decoy Pose()
:param working_Fc_glycan_chains: list( the chain id's for the working Fc glycan ). Ex = [ 'H', 'I' ]
:param native: native Pose()
:param native_Fc_glycan_chains: list( the chain id's for the native Fc glycan ). Ex = [ 'D', 'E' ]
:param decoy_num: int( the number of the decoy for use when dumping its Fc glycan )
:param dump_dir: str( /path/to/dump_dir for the temp pdb files made. Files will be deleted )
return: float( Fc glycan rmsd )
"""
# imports
import os
from rosetta import Pose
from rosetta.core.scoring import non_peptide_heavy_atom_RMSD
from antibody_functions import load_pose
from util import dump_pdb_by_chain, id_generator
# get temporary files to work with
id = id_generator()
if dump_dir.endswith( '/' ):
working_filename = "%s%s_temp_working_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
native_filename = "%s%s_temp_native_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
else:
working_filename = "%s/%s_temp_working_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
native_filename = "%s/%s_temp_native_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
# dump out the Fc glycans by their chain id's
dump_pdb_by_chain( working_filename, working, working_Fc_glycan_chains, decoy_num, dump_dir = dump_dir )
dump_pdb_by_chain( native_filename, native, native_Fc_glycan_chains, decoy_num, dump_dir = dump_dir )
# load in the Fc glycans
just_Fc_glycan = Pose()
try:
just_Fc_glycan.assign( load_pose( working_filename ) )
except:
pass
native_just_Fc_glycan = Pose()
try:
native_just_Fc_glycan.assign( load_pose( native_filename ) )
except:
pass
# calculate the glycan rmsd
try:
glycan_rmsd = non_peptide_heavy_atom_RMSD( just_Fc_glycan, native_just_Fc_glycan )
except:
glycan_rmsd = "nan"
pass
# delete the files
try:
os.popen( "rm %s" %working_filename )
os.popen( "rm %s" %native_filename )
except:
pass
return glycan_rmsd
def pose_from_pubchem(cid, name, temporary=True):
pose = Pose()
if temporary:
# the temporary solution, create an ephemeral ResidueSet
params_from_pubchem(cid, name)
# generate ResidueSet
res_set = generate_nonstandard_residue_set([name])
# fill the pose
pose_from_pdb(pose, res_set, name + '_0001.pdb')
else:
# permanent solution, add to .params list
add_cid_to_database(cid, name)
# fill the pose
pose_from_pdb(pose, name + '_0001.pdb')
return pose
for seq_pos in native_Fc_protein_nums:
res = native_pose.residue( seq_pos )
# used to store scores for each mutation to then determine which was best
AA_to_score_dict = {}
# skipping Cys residues because they're just being sassy
if not res.name1() == 'C':
# don't do sugars
if not res.is_carbohydrate():
# skip branch points like ASN 297 since it connects to GlcNAc 1
if not res.is_branch_point():
# for each amino acid
for new_AA in AA_name1_list:
# get a copy of the Pose
mutant = Pose()
mutant.assign( native_pose )
# mutate!
mutant.assign( mutate_residue( seq_pos, new_AA, mutant, sf, pack_radius = 10 ) )
# change pose name to (orig AA)(pdb number and chain)_to_newAA
orig_AA_name1 = str( native_pose.residue( seq_pos ).name1() )
mutant.pdb_info().name( "%s%s_to_%s" %( orig_AA_name1, str( mutant.pdb_info().pose2pdb( seq_pos ).strip().replace( ' ', '' ) ), new_AA ))
# visualize mutation
#pmm.apply(mutant)
# store all information
# store all data for this residue's mutation for the df
# append original residue, position, and score to list
# utility functions from file_mover_based_on_fasc import main as get_lowest_E_from_fasc from get_pose_metrics_on_native import main as get_pose_metrics_on_native from pose_metrics_util import Fc_glycan_rmsd ################################ #### INITIAL PROTOCOL SETUP #### ################################ # initialize Rosetta initialize_rosetta() # load up the poses given from the arguments passed native_pose = Pose() native_pose.assign( load_pose( input_args.native_pdb_file ) ) # automatically populates chain and residue information into holder for native 3ay4 native_pose_info = hold_chain_and_res_designations_3ay4() native_pose_info.native() # get the full path of the native PDB name native_pdb_filename_full_path = input_args.native_pdb_file native_pdb_filename = native_pdb_filename_full_path.split( '/' )[-1] native_pdb_name = native_pdb_filename.split( ".pdb" )[0] # change the name of the native PDB name native_pose_name = "native_pose" native_pose.pdb_info().name( native_pose_name )
n_res = initial_pose.total_residue()
for seq_pos in range(1, n_res + 1):
res = initial_pose.residue(seq_pos)
for AA in AAs:
if res.name1() != AA:
variant_name = \
res.name1() + \
str(initial_pose.pdb_info().number(seq_pos)) + \
AA
# Check for disulfide special case.
if res.name() == 'CYD':
disulfide_partner = res.residue_connection_partner(
res.n_possible_residue_connections())
temp_pose = Pose()
temp_pose.assign(initial_pose)
# (Packing causes seg fault if current CYS residue is not
# also converted before mutating.)
change_cys_state(seq_pos, 'CYS', temp_pose.conformation())
change_cys_state(disulfide_partner, 'CYS',
temp_pose.conformation())
# Mutate protein.
mutant_pose = mutate_residue(temp_pose, seq_pos, AA,
PACK_RADIUS, sf)
else:
# Mutate protein.
mutant_pose = mutate_residue(initial_pose, seq_pos, AA,
PACK_RADIUS, sf)
# Minimize.
packer = RotamerTrialsMover(sf, pt)
# Prepare data storage.
surface = []
params = []
if args.make_movie:
dir_name = args.pdb_filename[:-4] + '_movie_frames'
try:
mkdir(dir_name)
except OSError:
print 'Warning: Directory already exists; files will be overwritten.'
# Prepare temp pose.
pose = Pose()
# Switch based on ring size.
ring_size = info.ring_size()
if ring_size == 3:
exit(' Functionality for 3-membered rings not coded yet; exiting')
elif ring_size == 4:
exit(' Functionality for 4-membered rings not coded yet; exiting')
elif ring_size == 5:
exit(' Functionality for 5-membered rings not coded yet; exiting')
elif ring_size == 6:
# Generate header for data output.
header = "\t"
for phi in range(0, 360 + 1, 30):
header += str(phi)
if phi == 360:
orig_pdb_filename = orig_pdb_filename_full_path.split( '/' )[-1]
orig_pdb_name = orig_pdb_filename.split( ".pdb" )[0]
# this is where the double pack and minimized versions of the native will be dumped
structure_dir = base_structs_dir
decoy_pdb_name = structure_dir + orig_pdb_name
# this is where the single pack and minimized decoy will be dumped (the predecessor to double)
# see long comment in jd code for clarification
round1_structure_dir = base_structs_dir + "/round1_decoys/"
if not os.path.isdir( round1_structure_dir ):
os.mkdir( round1_structure_dir )
round1_decoy_pdb_name = round1_structure_dir + orig_pdb_name
# sets up the input native PDB as being the base pose
native_pose = Pose()
native_pose.assign( load_pose( orig_pdb_filename_full_path ) )
native_pose.pdb_info().name( "native" )
# automatically populates chain and residue information into holder for native 3ay4
native_pose_info = hold_chain_and_res_designations_3ay4()
native_pose_info.native()
# fa_intra_rep should always 0.440 since that's what I've been using for sugars
sf = get_fa_scorefxn()
sf.set_weight( fa_intra_rep, 0.440 )
orig_fa_rep = sf.get_weight( fa_rep )
# pymol stuff
def main():
#takes name of pdb file without the extention
args = sys.argv
pdb_file = args[1]
out_file = args[2]
score_type = int(args[3])
#set up timer to figure out how long the code took to run
t0=time()
# Initialize Rosetta.
init(extra_options='-mute basic -mute core -mute protocol -mute warn')
# Constants
PACK_RADIUS = 5
#Amino acids, notice there is no C
AAs = ("A","D","E","F","G","H","I","K","L","M","N","P","Q","R","S","T","V","W","Y")
#Number of mutations to accept
max_accept_mut = 2000
#Population size
N = 1
#Beta (temp term)
beta = 1
#Prepare data headers
data = ['Variant,ChainA,ChainB,ChainC,InterfaceAB,InterfaceAC,"delta-delta-G",Probability,Generation\n']
initial_pose = pose_from_pdb(pdb_file)
#Set up ScoreFunction
sf = get_fa_scorefxn()
#Set up MoveMap This is where you turn the bb and side chain flexibility on and off
mm = MoveMap()
mm.set_bb(False)
#Get the init score of the struct to calc the threshold
pre_pre_packing_score = sf(initial_pose)
print(pre_pre_packing_score)
min_mover = MinMover()
min_mover.movemap(mm)
min_mover.score_function(sf)
min_mover.min_type('dfpmin_armijo_nonmonotone')
cp_init_pdb = Pose()
cp_init_pdb.assign(initial_pose)
chains=cp_init_pdb.split_by_chain()
#split up AB inter and AC inter
initial_poseAB = Pose()
initial_poseAB.assign(initial_pose)
initial_poseAC = Pose()
initial_poseAC.assign(initial_pose)
init_chain_moverAB = SwitchChainOrderMover()
init_chain_moverAB.chain_order("12")
init_chain_moverAB.apply(initial_poseAB)
init_chain_moverAC = SwitchChainOrderMover()
init_chain_moverAC.chain_order("13")
init_chain_moverAC.apply(initial_poseAC)
#score the inital stabs of each chain
wt_a=sf(chains[1])
wt_b=sf(chains[2])
wt_c=sf(chains[3])
#score the intial interfaces
inter_AB=InterfaceEnergy_split(initial_poseAB)
inter_AC=InterfaceEnergy_split(initial_poseAC)
#init thresholds set to half of the init stabilities, if you want to do a different protein change these
threshold_a=-138.41754752
threshold_b=-61.378619136
threshold_c=-61.378619136
threshold_inter_ab=-10.3726691079
threshold_inter_ac=-10.3726691079
data.append('WT,' + str(wt_a)+','+str(wt_b)+','+str(wt_c)+','+str(inter_AB)+','+str(inter_AC)+',0.0,0.0,0\n')
#check the inital starting score
init_score=score_all(initial_pose,sf,min_mover,beta,threshold_a, threshold_b, threshold_c,threshold_inter_ab,threshold_inter_ac,score_type)
print(init_score)
#number of residues to select from
n_res = initial_pose.total_residue()
print(n_res)
#start sim
i=0
gen=0
while i < max_accept_mut:
#update the number of generations that have pased
gen+=1
print 'accepts:', i
#pick a place to mutate
mut_location = random.randint(1, n_res)
#mut_location = random.randint(1, 10)
#get the amino acid at that position
res = initial_pose.residue(mut_location)
#don't mess with C, just choose again
while(res.name1() == 'C'):
mut_location = random.randint(1, n_res)
#get the amino acid at that position
res = initial_pose.residue(mut_location)
#choose the amino acid to mutate to
toname = res.name1()
new_mut_key = random.randint(0,len(AAs)-1)
proposed_res = AAs[new_mut_key]
#don't bother mutating to the same amino acid it just takes more time
while(proposed_res == res.name1()):
new_mut_key = random.randint(0,len(AAs)-1)
proposed_res = AAs[new_mut_key]
#init mutant with current
mutant_pose = Pose()
mutant_pose.assign(initial_pose)
#mutate
mutant_pose=mutate_residue_chain(mutant_pose, mut_location, proposed_res, PACK_RADIUS, sf)
#score mutant
mut_score=score_all(mutant_pose,sf,min_mover,beta,threshold_a, threshold_b, threshold_c,threshold_inter_ab,threshold_inter_ac,score_type)
#get the probability that the mutation will be accepted
probability = calc_prob_scores(mut_score['score'], init_score['score'], N)
rand = random.random()
#test to see if mutation is accepted
if float(rand) < float(probability):
print "accepted"
#make a name for the new mutant
variant_name = str(toname) + str(initial_pose.pdb_info().number(mut_location)) + str(proposed_res)
# Assuming some burn in phase, make this zero if you want to store everything
if i>=0:
#save name and energy change
data.append(variant_name +',' + str(mut_score['a'])+','+str(mut_score['b'])+','+str(mut_score['c'])+','+str(mut_score['ab'])+','+str(mut_score['ac'])+',' + str(mut_score['score'] - init_score['score']) + "," + str(probability) + "," + str(gen) + "\n")
#save the new accepted mutation
pdb_name=str(i)+".pdb"
mutant_pose.dump_pdb(pdb_name)
#update the wildtype
initial_pose = mutant_pose
init_score = mut_score
#update number of accepts
i+=1
def score_all(pdb,sf,min_mover,beta,threshold_a, threshold_b,threshold_inter_ab,score_type):
cp_init_pdb = Pose()
cp_init_pdb.assign(pdb)
chains = cp_init_pdb.split_by_chain()
initial_poseAB = Pose()
initial_poseAB.assign(pdb)
initial_poseAC = Pose()
initial_poseAC.assign(pdb)
init_chain_moverAB = SwitchChainOrderMover()
init_chain_moverAB.chain_order("12")
init_chain_moverAB.apply(initial_poseAB)
init_a=sf(chains[1])
init_b=sf(chains[2])
init_inter_AB=InterfaceEnergy_split_init(initial_poseAB)
data=[init_a,init_b,init_inter_AB]
thresh=[threshold_a,threshold_b,threshold_inter_ab]
init_score=0
if score_type == 1:
init_score=calc_x_list(data,beta,thresh)
if score_type == 2:
init_score=calc_x_list_2(data,beta,thresh)
if score_type == 3:
init_score=calc_x_list_3(data,beta,thresh)
if score_type == 4:
data=[init_a,init_b,init_c]
thresh=[threshold_a,threshold_b,threshold_c]
init_score=calc_x_list(data,beta,thresh)
dic = {'score':init_score, 'a':init_a, 'b':init_b, 'ab':init_inter_AB}
return(dic)
if f.endswith( ".pdb" ):
structures.append( os.path.abspath( f ) )
structure_names.append( f.split( '/' )[-1] )
os.chdir( working_dir )
# inform the user of the structure directory and number of files to be analyzed
num_structs = len( structure_names )
print "Analyzing", num_structs, "structures from", structure_dir
print
# check and load native pose
initialize_rosetta()
try:
if os.path.isfile( input_args.native_pdb_filename ):
native = Pose()
native.assign( load_pose( input_args.native_pdb_filename ) )
except:
print "It appears", input_args.native_pdb_filename, "is not a valid pdb file. Exiting"
sys.exit()
# make a scorefunction
sf = get_fa_scorefxn()
# collect the data
rmsds = []
scores = []
pdb_names = []
# for if Pandas doesn't work
def pose_from_params(filename, params_list):
res_set = generate_nonstandard_residue_set(params_list)
pose = Pose()
pose_from_pdb(pose, res_set, filename)
return pose
def Fc_glycan_metrics( working, native, working_Fc_glycan_chains, native_Fc_glycan_chains, sf, decoy_num, dump_dir ):
"""
Return the glycan RMSD contribution of the two Fc glycans in 3ay4 (may work for other PDBs, but I don't know yet)
Fc_glycan_buried_sasa = complex with Fc glycan - ( complex without Fc glycan + just Fc glycan )
hbonds contributed by Fc glycans = total hbonds in Pose - total hbonds in Pose without Fc glycans - just Fc glycan hbonds
:param working: decoy Pose()
:param native: native Pose()
:param working_Fc_glycan_chains: list( the chain id's for the working Fc glycan ). Ex = [ 'H', 'I' ]
:param native_Fc_glycan_chains: list( the chain id's for the native Fc glycan ). Ex = [ 'D', 'E' ]
:param sf: ScoreFunction
:param decoy_num: int( the number of the decoy for use when dumping its Fc glycan )
:param dump_dir: str( /path/to/dump_dir for the temp pdb files made. Files will be deleted )
:return: obj( DataHolder that contains Fc_glycan_rmsd, Fc_glycan_tot_score, Fc_glycan_buried_sasa, and Fc_glycan_internal_hbonds, Fc_glycan_hbonds_contributed )
"""
#################
#### IMPORTS ####
#################
# Rosetta functions
from rosetta import Pose, calc_total_sasa
from rosetta.core.scoring import non_peptide_heavy_atom_RMSD
# Rosetta functions I wrote out
from antibody_functions import load_pose, DataHolder
# utility functions
import os
from util import dump_pdb_by_chain, id_generator
from toolbox import get_hbonds
# for use in SASA calculations
probe_size = 1.4
# get glycan rmsd (not using above function because I want to use the glycan poses for something else
# get temporary files to work with
id = id_generator()
if dump_dir.endswith( '/' ):
working_filename = "%s%s_temp_working_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
native_filename = "%s%s_temp_native_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
else:
working_filename = "%s/%s_temp_working_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
native_filename = "%s/%s_temp_native_just_glyc%s.pdb" %( dump_dir, id, str( decoy_num ) )
# dump out the Fc glycans by their chain id's
dump_pdb_by_chain( working_filename, working, working_Fc_glycan_chains, decoy_num, dump_dir = dump_dir )
dump_pdb_by_chain( native_filename, native, native_Fc_glycan_chains, decoy_num, dump_dir = dump_dir )
# load in the Fc glycans
working_just_Fc_glycan = Pose()
try:
working_just_Fc_glycan.assign( load_pose( working_filename ) )
except:
pass
native_just_Fc_glycan = Pose()
try:
native_just_Fc_glycan.assign( load_pose( native_filename ) )
except:
pass
# calculate the glycan rmsd
try:
glycan_rmsd = non_peptide_heavy_atom_RMSD( working_just_Fc_glycan, native_just_Fc_glycan )
except:
glycan_rmsd = "nan"
pass
# get the metrics associated with just the Fc glycan
# score first as to gain access to the hbonds data
working_Fc_glycan_tot_score = sf( working_just_Fc_glycan )
native_Fc_glycan_tot_score = sf( native_just_Fc_glycan )
# SASA of just the glycan
working_Fc_glycan_sasa = calc_total_sasa( working_just_Fc_glycan, probe_size )
native_Fc_glycan_sasa = calc_total_sasa( native_just_Fc_glycan, probe_size )
# num hbonds in Fc glycan
working_Fc_glycan_internal_hbonds = get_hbonds( working_just_Fc_glycan ).nhbonds()
native_Fc_glycan_internal_hbonds = get_hbonds( native_just_Fc_glycan ).nhbonds()
# delete the files
try:
os.popen( "rm %s" %working_filename )
os.popen( "rm %s" %native_filename )
except:
pass
# now move to metrics requiring the removal of the glycan from the complex
# get temporary files to work with
id = id_generator()
if dump_dir.endswith( '/' ):
working_filename = "%s%s_working_no_glyc_%s.pdb" %( dump_dir, id, str( decoy_num ) )
native_filename = "%s%s_native_no_glyc_%s.pdb" %( dump_dir, id, str( decoy_num ) )
else:
working_filename = "%s/%s_working_no_glyc_%s.pdb" %( dump_dir, id, str( decoy_num ) )
native_filename = "%s/%s_native_no_glyc_%s.pdb" %( dump_dir, id, str( decoy_num ) )
# get the chain id's of everything discluding the passed Fc glycan chain id's
working_pose_chains = []
for res in working:
chain_id = working.pdb_info().chain( res.seqpos() )
if ( chain_id not in working_pose_chains ) and ( chain_id not in working_Fc_glycan_chains ):
working_pose_chains.append( chain_id )
native_pose_chains = []
for res in native:
chain_id = native.pdb_info().chain( res.seqpos() )
if ( chain_id not in native_pose_chains ) and ( chain_id not in native_Fc_glycan_chains ):
native_pose_chains.append( chain_id )
# dump out the pose without its Fc glycans by the chain id's
dump_pdb_by_chain( working_filename, working, working_pose_chains, decoy_num, dump_dir = dump_dir )
dump_pdb_by_chain( native_filename, native, native_pose_chains, decoy_num, dump_dir = dump_dir )
# load in the working Pose without the Fc glycans
working_complex_no_Fc_glycan = Pose()
native_complex_no_Fc_glycan = Pose()
try:
working_complex_no_Fc_glycan.assign( load_pose( working_filename ) )
native_complex_no_Fc_glycan.assign( load_pose( working_filename ) )
except:
pass
# score the Poses so their hbond energies get updated
sf( working )
sf( working_complex_no_Fc_glycan )
sf( native )
sf( native_complex_no_Fc_glycan )
# get the number of hbonds in the Pose without the Fc glycans
# working
working_with_Fc_glycan_hbonds = get_hbonds( working )
working_no_Fc_glycan_hbonds = get_hbonds( working_complex_no_Fc_glycan )
working_Fc_glycan_hbonds_contributed = working_with_Fc_glycan_hbonds.nhbonds() - working_no_Fc_glycan_hbonds.nhbonds() - working_Fc_glycan_internal_hbonds
# native
native_with_Fc_glycan_hbonds = get_hbonds( native )
native_no_Fc_glycan_hbonds = get_hbonds( native_complex_no_Fc_glycan )
native_Fc_glycan_hbonds_contributed = native_with_Fc_glycan_hbonds.nhbonds() - native_no_Fc_glycan_hbonds.nhbonds() - native_Fc_glycan_internal_hbonds
# get the SASA contributed by the presence of the Fc glycan
# working
working_with_Fc_glycan_sasa = calc_total_sasa( working, probe_size )
working_no_Fc_glycan_sasa = calc_total_sasa( working_complex_no_Fc_glycan, probe_size )
working_Fc_glycan_sasa_contributed = working_with_Fc_glycan_sasa - ( working_no_Fc_glycan_sasa + working_Fc_glycan_sasa )
# native
native_with_Fc_glycan_sasa = calc_total_sasa( native, probe_size )
native_no_Fc_glycan_sasa = calc_total_sasa( native_complex_no_Fc_glycan, probe_size )
native_Fc_glycan_sasa_contributed = native_with_Fc_glycan_sasa - ( native_no_Fc_glycan_sasa + native_Fc_glycan_sasa )
# delete the files
try:
os.popen( "rm %s" %working_filename )
os.popen( "rm %s" %native_filename )
except:
pass
# store data in the DataHolder and return it
data = DataHolder()
data.Fc_glycan_rmsd = glycan_rmsd
data.Fc_glycan_tot_score = working_Fc_glycan_tot_score
data.native_Fc_glycan_tot_score = native_Fc_glycan_tot_score
data.Fc_glycan_internal_hbonds = working_Fc_glycan_internal_hbonds
data.native_Fc_glycan_internal_hbonds = native_Fc_glycan_internal_hbonds
data.Fc_glycan_hbonds_contributed = working_Fc_glycan_hbonds_contributed
data.native_Fc_glycan_hbonds_contributed = native_Fc_glycan_hbonds_contributed
data.Fc_glycan_sasa_contributed = working_Fc_glycan_sasa_contributed
data.native_Fc_glycan_sasa_contributed = native_Fc_glycan_sasa_contributed
data.probe_size = probe_size
return data
# keeping this here for now in case I decide to do the alanine scan on the whole protein
if seq_pos in native_Fc_protein_nums:
# use the contact map created earlier to determine if this residue is near the glycan
# since the contact map was made between the protein and glycan, just check to see if this res is a key
within_10A_of_Fc_glycan_res.append( seq_pos in Fc_protein_to_Fc_glycan_cmap.keys() )
# use the contact map to determine if this residue is within 10A of the Fc-FcR interface
within_10A_of_FcR_interface.append( seq_pos in Fc_protein_to_FcR_protein_cmap.keys() )
# otherwise it's a residue outside the Fc protein, so just put None
else:
within_10A_of_Fc_glycan_res.append( None )
# get a copy of the Pose
mutant = Pose()
mutant.assign( native_pose )
# build the new Ala residue based on the position of the old residue
# if this is the first residue of the pose
if seq_pos == 1:
# this is the N-terminal end, so use an N-terminal ALA residue
ALA_restype = ALA_pose.conformation().residue_type( 1 )
# if this is the last residue of the pose
elif seq_pos == native_pose.n_residue():
# this is the C-terminal end, so use an C-terminal ALA residue
ALA_restype = ALA_pose.conformation().residue_type( 3 )
# otherwise, this is a residue with something bonded on both sides
else:
def pseudo_interface_energy_3ay4( pose, in_sf, native = False, pmm = None ):
"""
Attempts to get pseudo-interface energy of a glycosylated 3ay4 decoy
Lots of hard coding here - works on a decoy pose as Rosetta renumbers the Pose a bit
Makes the two ASN connections to the Fc A and B glycans JUMPs instead of chemical EDGEs
:param pose: Pose
:param in_sf: ScoreFunction
:param native: bool( is this the native 3ay4 or a decoy? Answer determines how FoldTree gets coded )
:param pmm: PyMOL_Mover( pass a PyMOL_Mover object if you want to watch the protocol ). Default = None
:return: float( pseudo interface energy )
"""
from rosetta import FoldTree, Pose
from rosetta.numeric import xyzVector_Real
from rosetta.core.scoring import score_type_from_name
# if this isn't the Fc-FcR structure of 3ay4, just return 0
if pose.n_residue() != 618:
return 0
# set atom_pair_constraint weight to 0
sf = in_sf.clone()
sf.set_weight( score_type_from_name( "atom_pair_constraint" ), 0.0 )
# get the score of the whole complex
start_score = sf( pose )
# hard code the new FoldTree specific to a glycosylated decoy of 3ay4
if not native:
ft = FoldTree()
ft.add_edge( 1, 215, -1 )
ft.add_edge( 1, 216, 1 ) # beginning of chain A to beginning of chain B
ft.add_edge( 216, 431, -1 )
ft.add_edge( 1, 432, 2 ) # beginning of chain A to beginning of chain C
ft.add_edge( 432, 591, -1 )
ft.add_edge( 579, 592, "ND2", "C1" )
ft.add_edge( 592, 596, -1 )
ft.add_edge( 594, 597, "O6", "C1" )
ft.add_edge( 597, 598, -1 )
ft.add_edge( 592, 599, "O6", "C1" )
ft.add_edge( 462, 600, "ND2", "C1" )
ft.add_edge( 600, 602, -1 )
ft.add_edge( 69, 603, 3 ) # ASN 297 A to core GlcNAc H
ft.add_edge( 603, 607, -1 )
ft.add_edge( 605, 608, "O6", "C1" )
ft.add_edge( 608, 610, -1 )
ft.add_edge( 284, 611, 4 ) # ASN 297 B to core GlcNAc J
ft.add_edge( 611, 615, -1 )
ft.add_edge( 613, 616, "O6", "C1" )
ft.add_edge( 616, 618, -1 )
# hard code the new FoldTree specific to the native 3ay4
else:
ft = FoldTree()
ft.add_edge( 1, 215, -1 )
ft.add_edge( 69, 216, 1 ) # ASN 297 A to core GlcNAc D
ft.add_edge( 216, 220, -1 )
ft.add_edge( 218, 221, "O6", "C1" )
ft.add_edge( 221, 223, -1 )
ft.add_edge( 1, 224, 2 ) # beginning of chain A to beginning of chain B
ft.add_edge( 224, 439, -1 )
ft.add_edge( 292, 440, 3 ) # ASN 297 B to core GlcNAc E
ft.add_edge( 440, 444, -1 )
ft.add_edge( 442, 445, "O6", "C1" )
ft.add_edge( 445, 447, -1 )
ft.add_edge( 1, 448, 4 ) # beginning of chain A to beginning of chain C
ft.add_edge( 448, 607, -1 )
ft.add_edge( 595, 608, "ND2", "C1" )
ft.add_edge( 608, 612, -1 )
ft.add_edge( 610, 613, "O6", "C1" )
ft.add_edge( 613, 614, -1 )
ft.add_edge( 608, 615, "O6", "C1" )
ft.add_edge( 478, 616, "ND2", "C1" )
ft.add_edge( 616, 618, -1 )
# make a temporary Pose and give the new FoldTree to it
try:
pmm.keep_history( True )
except:
pass
temp_pose = Pose()
temp_pose.assign( pose )
temp_pose.fold_tree( ft )
try:
pmm.apply( temp_pose )
except:
pass
# split apart the two Fc sugars one-by-one
# if decoy structure -- the two glycans are now the last two new jumps
if not native:
jump = temp_pose.jump( 3 ) # sugar A
vec = xyzVector_Real( 1000, 1000, 1000 )
jump.set_translation( vec )
temp_pose.set_jump( 3, jump )
try:
pmm.apply( temp_pose )
except:
pass
jump = temp_pose.jump( 4 ) # sugar B
vec = xyzVector_Real( 1000, 1000, 1000 )
jump.set_translation( vec )
temp_pose.set_jump( 4, jump )
try:
pmm.apply( temp_pose )
except:
pass
# else native structure -- the two glycans are the first and third new jumps
else:
jump = temp_pose.jump( 1 ) # sugar A
vec = xyzVector_Real( 1000, 1000, 1000 )
jump.set_translation( vec )
temp_pose.set_jump( 1, jump )
try:
pmm.apply( temp_pose )
except:
pass
jump = temp_pose.jump( 3 ) # sugar B
vec = xyzVector_Real( 1000, 1000, 1000 )
jump.set_translation( vec )
temp_pose.set_jump( 3, jump )
try:
pmm.apply( temp_pose )
except:
pass
# score the split-apart Pose
split_score = sf( temp_pose )
# get the pseudo-interface score
# total - split = interface ( ie. interface + split = total )
pseudo_interface_energy = start_score - split_score
return pseudo_interface_energy
# Prepack.
packer.apply(pose)
visualize(pose)
print ' Score after prepacking:', sf(pose)
sf.show(pose) # TEMP
# Fold the polysaccharide.
print 'Folding/Refining structural polysaccharide with',
print pose.total_residue(),
#print pose.residue(2).carbohydrate_info().short_name(), 'residues...'
print 'residues...'
# Prepare job distributor.
jd = PyJobDistributor(args.output_folder + output_base_filename,
args.n_decoys, sf)
starting_pose = Pose()
starting_pose.assign(pose)
while not jd.job_complete:
print ' Decoy', jd.current_num
print ' Randomizing positions...'
pose.assign(starting_pose)
# Run protocol.
print ' Refining...'
refine_saccharide(pose, args)
jd.output_decoy(pose)
args.make_movie = False # Make no more than one movie.
# utility functions from file_mover_based_on_fasc import main as get_lowest_E_from_fasc from get_pose_metrics_on_native import main as get_pose_metrics_on_native from pose_metrics_util import Fc_glycan_rmsd ################################ #### INITIAL PROTOCOL SETUP #### ################################ # initialize Rosetta initialize_rosetta() # load up the poses given from the arguments passed native_pose = Pose() native_pose.assign( load_pose( input_args.native_pdb_file ) ) # automatically populates chain and residue information into holder for native 3ay4 native_pose_info = hold_chain_and_res_designations_3ay4() native_pose_info.native() # get the full path of the native PDB name native_pdb_filename_full_path = input_args.native_pdb_file native_pdb_filename = native_pdb_filename_full_path.split( '/' )[-1] native_pdb_name = native_pdb_filename.split( ".pdb" )[0] # change the name of the native PDB name native_pose_name = "native_pose" native_pose.pdb_info().name( native_pose_name )
n_res = initial_pose.total_residue()
for seq_pos in range(1, n_res + 1):
res = initial_pose.residue(seq_pos)
for AA in AAs:
if res.name1() != AA:
variant_name = \
res.name1() + \
str(initial_pose.pdb_info().number(seq_pos)) + \
AA
# Check for disulfide special case.
if res.name() == 'CYD':
disulfide_partner = res.residue_connection_partner(
res.n_residue_connections())
temp_pose = Pose()
temp_pose.assign(initial_pose)
# (Packing causes seg fault if current CYS residue is not
# also converted before mutating.)
change_cys_state(seq_pos, 'CYS',
temp_pose.conformation())
change_cys_state(disulfide_partner, 'CYS',
temp_pose.conformation())
# Mutate protein.
mutant_pose = mutate_residue(temp_pose, seq_pos, AA,
PACK_RADIUS, sf)
else:
# Mutate protein.
mutant_pose = mutate_residue(initial_pose, seq_pos, AA,
PACK_RADIUS, sf)
# utility functions from file_mover_based_on_fasc import main as get_lowest_E_from_fasc from get_pose_metrics import main as get_pose_metrics ################################ #### INITIAL PROTOCOL SETUP #### ################################ # initialize Rosetta initialize_rosetta() ## load up the poses given from the arguments passed # native pose ( for comparison, really ) native_pose = Pose() native_pose.assign( load_pose( input_args.native_pdb_file ) ) # get the full path of the native PDB name native_pdb_filename_full_path = input_args.native_pdb_file native_pdb_filename = native_pdb_filename_full_path.split( '/' )[-1] native_pdb_name = native_pdb_filename.split( ".pdb" )[0] # change the name of the native PDB name native_pose_name = "native_pose" native_pose.pdb_info().name( native_pose_name ) # load up the working pose working_pose = Pose() working_pose.assign( load_pose( input_args.working_pdb_file ) )
jump_minimizer = MinMover(mm, sf, 'dfpmin', 0.01, True)
#dock_hires = DockMCMProtocol() # is not rigid-body
#dock_hires.set_scorefxn(sf)
#dock_hires.set_partners(partners)
# Prepare job distributor.
jd = PyJobDistributor(new_filename[:-4], args.n_decoys, sf)
if args.ref:
if not args.ref.endswith('.pdb'):
exit('Reference file must have the ".pdb" file extension.')
ref_pose = pose_from_pdb(args.ref)
# Begin docking protocol.
print '\nDocking...'
pose = Pose() # working pose
last_pose = Pose() # needed because I have not modified C++ code
while not jd.job_complete:
print ' Decoy', jd.current_num
print ' Randomizing positions...'
pose.assign(starting_pose)
if not args.local:
randomizerA.apply(pose)
randomizerB.apply(pose)
visualize(pose)
slider.apply(pose)
visualize(pose)
print ' Ligand center is',
print pose.jump(JUMP_NUM).get_translation().length,
print 'Angstroms from protein center.'
pdb_chain = 'A',
do_pack = False,
do_min = False ) )
# mutate chain B
mutant_pose.assign( mutate_residue( pdb_num, new_residue, mutant_pose, sf,
pdb_num = True,
pdb_chain = 'B',
do_pack = False,
do_min = False ) )
except:
continue
###########################
#### gradient pack/min ####
###########################
min_pose = Pose()
for ii in range( 1 ):
working_pose = mutant_pose.clone()
for jj in range( 1 ):
# packing
task = standard_packer_task( working_pose )
task.or_include_current( True )
task.restrict_to_repacking()
rtm = RotamerTrialsMover( sf, task )
rtm.apply( working_pose )
# minimizing
mm = MoveMap()
mm.set_bb( True )
mm.set_chi( True )
def sequence_mapping(pdb_file, sequence_file, score_file, relax, jobs):
if os.path.exists( os.getcwd() + '/' + pdb_file ) and pdb_file:
init()
pose = Pose()
score_fxn = create_score_function('talaris2014')
if (relax):
refinement = FastRelax(score_fxn)
pose_from_pdb(pose, pdb_file)
if os.path.exists( os.getcwd() + '/' + sequence_file ) and sequence_file:
fid = open(sequence_file,'r')
fod = open(score_file,'w')
data = fid.readlines()
fid.close()
sequences = []
read_seq = False
for i in data:
if not len(i):
continue
elif i[0] == '>':
read_seq = True
fasta_line = re.split(':|\s+|\||\\n',i[1:])
name_cpt=0
while (name_cpt<len(fasta_line) and not fasta_line[name_cpt]):
name_cpt+=1
if name_cpt<len(fasta_line):
job_output = fasta_line[name_cpt]
else:
print 'Error: Please enter an identifier for sequences in your fasta file'
exit(1)
elif read_seq:
seq=list(i)
resn=1
for j in i:
if j!='\n' and resn<=pose.total_residue():
mutator = MutateResidue( resn , one_to_three[j] )
mutator.apply( pose )
resn+=1
elif resn>pose.total_residue():
print 'WARNING: couldn\'t mutate residue number '+str(resn)+', sequence too long for backbone...'
resn+=1
if (relax):
jd = PyJobDistributor(job_output, jobs, score_fxn)
jd.native_pose = pose
scores = [0]*(jobs)
counter = 0
decoy=Pose()
while not jd.job_complete:
decoy.assign(pose)
resn=1
refinement.apply(decoy)
jd.output_decoy(decoy)
scores[counter]=score_fxn(decoy)
counter+=1
for i in range(0, len(scores)):
fod.writelines(job_output + '_' + str(i+1) + ' : '+str(scores[i])+'\n')
else:
pose_packer = standard_packer_task(pose)
pose_packer.restrict_to_repacking()
packmover = PackRotamersMover(score_fxn, pose_packer)
packmover.apply(pose)
fod.writelines(job_output+' : '+str(score_fxn(pose))+'\n')
pose.dump_pdb(job_output+'_1.pdb')
else:
print 'Bad fasta format'
exit(1)
fod.close()
else:
print 'Please provide a valid sequence file, '+sequence_file+' doesn\'t exist'
else:
print 'Please provide a valid backbone file, '+pdb_file+' doesn\'t exist'
# utility functions from file_mover_based_on_fasc import main as get_lowest_E_from_fasc from get_pose_metrics_on_native import main as get_pose_metrics_on_native from pose_metrics_util import Fc_glycan_rmsd ################################ #### INITIAL PROTOCOL SETUP #### ################################ # initialize Rosetta initialize_rosetta() # load up the poses given from the arguments passed native_pose = Pose() native_pose.assign( load_pose( input_args.native_pdb_file ) ) # automatically populates chain and residue information into holder for native 3ay4 native_pose_info = hold_chain_and_res_designations_3ay4() native_pose_info.native() # get the full path of the native PDB name native_pdb_filename_full_path = input_args.native_pdb_file native_pdb_filename = native_pdb_filename_full_path.split( '/' )[-1] native_pdb_name = native_pdb_filename.split( ".pdb" )[0] # change the name of the native PDB name native_pose_name = "native_pose" native_pose.pdb_info().name( native_pose_name )
def pose_from_sequence(seq, res_type='fa_standard', name='', chain_id='A'):
"""
Returns a pose generated from amino acid single letters in <seq> using
the <res_type> ResidueType, the new pose's PDBInfo is named <name>
and all residues have chain ID <chain_id>
example:
pose=pose_from_sequence('LIGAND')
See also:
Pose
make_pose_from_sequence
pose_from_pdb
pose_from_rcsb
"""
pose = Pose()
make_pose_from_sequence(pose, seq, res_type)
#pdb_info = rosetta.core.pose.PDBInfo(pose.total_residue()) # actual, for other code
pdb_info = PDBInfo(pose.total_residue()) # create a PDBInfo object
for i in range(0, pose.total_residue()):
if pose.residue(i + 1).is_protein():
# set to a more reasonable default
pose.set_phi(i + 1, -150)
pose.set_psi(i + 1, 150)
pose.set_omega(i + 1, 180)
# set PDBInfo info for chain and number
#pdb_info.chain(i+1,chain_id)
#pdb_info.number(i+1,i+1)
#### you can alternatively use the deprecated method set_extended_torsions
#### which requires a Pose and a Loop object...so make a large loop
#set_extended_torsions( pose , Loop ( 1 , pose.total_residue() ) )
# set the PDBInfo
pose.pdb_info(pdb_info)
# default name to first 3 letters
if not name:
name = seq[:4]
pose.pdb_info().name(name)
# print pose
return pose
def score_all(pdb,sf,min_mover,beta,threshold_a, threshold_b, threshold_c,threshold_inter_ab,threshold_inter_ac,score_type):
cp_init_pdb = Pose()
cp_init_pdb.assign(pdb)
chains = cp_init_pdb.split_by_chain()
initial_poseAB = Pose()
initial_poseAB.assign(pdb)
initial_poseAC = Pose()
initial_poseAC.assign(pdb)
init_chain_moverAB = SwitchChainOrderMover()
init_chain_moverAB.chain_order("12")
init_chain_moverAB.apply(initial_poseAB)
init_chain_moverAC = SwitchChainOrderMover()
init_chain_moverAC.chain_order("13")
init_chain_moverAC.apply(initial_poseAC)
init_a=sf(chains[1])
init_b=sf(chains[2])
init_c=sf(chains[3])
init_inter_AB=InterfaceEnergy_split(initial_poseAB)
init_inter_AC=InterfaceEnergy_split(initial_poseAC)
data=[init_a,init_b,init_c,init_inter_AB,init_inter_AC]
thresh=[threshold_a,threshold_b,threshold_c,threshold_inter_ab,threshold_inter_ac]
init_score=0
#score everything (bind both)
if score_type == 1:
init_score=calc_x_list(data,beta,thresh)
#pentalize binding B' (bind B and not B')
if score_type == 2:
init_score=calc_x_list_2(data,beta,thresh)
#bind just the best duplicate (bind max)
if score_type == 3:
init_score=calc_x_list_3(data,beta,thresh)
#just stabilitys and no binding (no bind)
if score_type == 4:
data=[init_a,init_b,init_c]
thresh=[threshold_a,threshold_b,threshold_c]
init_score=calc_x_list(data,beta,thresh)
def mutate_residue(pose,
mutant_position,
mutant_aa,
pack_radius=0.0,
pack_scorefxn=''):
"""
Replaces the residue at <mutant_position> in <pose> with <mutant_aa>
and repack any residues within <pack_radius> Angstroms of the mutating
residue's center (nbr_atom) using <pack_scorefxn>
note: <mutant_aa> is the single letter name for the desired ResidueType
example:
mutate_residue(pose,30,A)
See also:
Pose
PackRotamersMover
MutateResidue
pose_from_sequence
"""
#### a MutateResidue Mover exists similar to this except it does not pack
#### the area around the mutant residue (no pack_radius feature)
#mutator = MutateResidue( mutant_position , mutant_aa )
#mutator.apply( test_pose )
if pose.is_fullatom() == False:
IOError('mutate_residue only works with fullatom poses')
test_pose = Pose()
test_pose.assign(pose)
# create a standard scorefxn by default
if not pack_scorefxn:
pack_scorefxn = create_score_function('standard')
task = standard_packer_task(test_pose)
# the Vector1 of booleans (a specific object) is needed for specifying the
# mutation, this demonstrates another more direct method of setting
# PackerTask options for design
aa_bool = rosetta.utility.vector1_bool()
# PyRosetta uses several ways of tracking amino acids (ResidueTypes)
# the numbers 1-20 correspond individually to the 20 proteogenic amino acids
# aa_from_oneletter returns the integer representation of an amino acid
# from its one letter code
# convert mutant_aa to its integer representation
mutant_aa = aa_from_oneletter_code(mutant_aa)
# mutation is performed by using a PackerTask with only the mutant
# amino acid available during design
# to do this, construct a Vector1 of booleans indicating which amino acid
# (by its numerical designation, see above) to allow
for i in range(1, 21):
# in Python, logical expression are evaluated with priority, thus the
# line below appends to aa_bool the truth (True or False) of the
# statement i == mutant_aa
aa_bool.append(i == mutant_aa)
# modify the mutating residue's assignment in the PackerTask using the
# Vector1 of booleans across the proteogenic amino acids
task.nonconst_residue_task(mutant_position).restrict_absent_canonical_aas(
aa_bool)
# prevent residues from packing by setting the per-residue "options" of
# the PackerTask
center = pose.residue(mutant_position).nbr_atom_xyz()
for i in range(1, pose.total_residue() + 1):
# only pack the mutating residue and any within the pack_radius
if not i == mutant_position or center.distance_squared(
test_pose.residue(i).nbr_atom_xyz()) > pack_radius**2:
task.nonconst_residue_task(i).prevent_repacking()
# apply the mutation and pack nearby residues
packer = PackRotamersMover(pack_scorefxn, task)
packer.apply(test_pose)
return test_pose
initialize_rosetta()
# get the full path to the original native PDB filename
orig_pdb_filename_full_path = input_args.native_pdb_file
orig_pdb_filename = orig_pdb_filename_full_path.split( '/' )[-1]
orig_pdb_name = orig_pdb_filename.split( ".pdb" )[0]
# make the directory for the native PDB in the base_structs_dir
# this is where packed and minimized versions of the native will lie
structure_dir = base_structs_dir + orig_pdb_name
if not os.path.isdir( structure_dir ):
os.mkdir( structure_dir )
decoy_pdb_name = structure_dir + '/' + orig_pdb_name
# sets up the input native PDB as being the base pose
native_pose = Pose()
native_pose.assign( load_pose( orig_pdb_filename_full_path ) )
native_pose.pdb_info().name( "native" )
# automatically populates chain and residue information into holder for native 3ay4
native_pose_info = hold_chain_and_res_designations_3ay4()
native_pose_info.native()
# use the scorefxn_file to set up additional weights
if input_args.scorefxn_file is not None:
sf = make_fa_scorefxn_from_file( input_args.scorefxn_file )
# else create a fa_scorefxn
else:
sf = get_fa_scorefxn()
initialize_rosetta()
pmm = PyMOL_Mover()
pmm.keep_history( True )
#working = load_pose( "just_native_sugar.pdb" )
native = load_pose( "/Users/Research/pyrosetta_dir/project_structs/lowest_E_double_pack_and_min_only_native_crystal_struct_3ay4_Fc_FcgRIII.pdb" )
native.pdb_info().name( "native" )
#pmm.apply( native )
#working = load_pose( "just_native_ASN.pdb" )
working = load_pose( "/Users/Research/pyrosetta_dir/project_structs/lowest_E_double_pack_and_min_only_native_crystal_struct_3ay4_Fc_FcgRIII_removed_Fc_sugar.pdb" )
working.pdb_info().name( "working" )
decoy = Pose()
decoy.assign( working )
decoy.pdb_info().name( "decoy" )
#pmm.apply( decoy )
glycosylate_these_ASN = [ 69, 284 ]
for ASN in glycosylate_these_ASN:
glycosylate_pose_by_file( decoy, ASN, "ND2", "/Users/Research/antibody_project/send_to_louis/project_glyco_files/3ay4_Fc_Glycan.iupac" )
decoy.pdb_info().name( "decoy" )
#pmm.apply( decoy )
for decoy_res in glycosylate_these_ASN:
native_res = decoy_to_native_res_map[ decoy_res ]
print decoy_res, native_res