def main():
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
logger.info(pcl.welcome())
def main():
starttime = time.time()
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
logger.info(pcl.welcome())
# READ INPUT ARGUMENTS
if args.initialise is not None:
inseq = pio.check_path(args.initialise[0], 'file')
instr = pio.check_path(args.initialise[1], 'file')
indb = pio.check_path(pio.conf.CSV_CHAIN_PATH, 'file')
skipexec = [False, True] if not args.add_noncropped else [False, False]
scoring = [True, False] if not args.add_noncropped else [True, True]
elif args.skip_conpred is not None:
inseq = pio.check_path(args.skip_conpred[0], 'file')
instr = pio.check_path(args.skip_conpred[1], 'file')
skipexec = [True, True]
scoring = [True, False] if not args.add_noncropped else [True, True]
elif args.skip_default_conpred is not None:
inseq = pio.check_path(args.skip_default_conpred[0], 'file')
instr = pio.check_path(args.skip_default_conpred[1], 'file')
skipexec = [True, True] if not args.add_noncropped else [True, False]
scoring = [True, False] if not args.add_noncropped else [True, True]
if args.outdir is None:
outrootdir = pio.check_path(os.path.dirname(inseq))
else:
outrootdir = pio.check_path(os.path.join(args.outdir[0], ''))
pio.mdir(outrootdir)
if args.collection_file is None:
outcsvfile = pio.check_path(
os.path.join(outrootdir, "pisacov_data.csv"))
else:
outcsvfile = pio.check_path(args.collection_file[0])
try:
pio.check_path(outcsvfile, 'file')
csvexists = True
except:
csvexists = False
if args.uniprot_threshold is not None:
thuprot, dbuprot = pio.check_uniprot(args.uniprot_threshold[0])
else:
thuprot = 0.0
dbuprot = None
if args.hhparams is not None:
hhparameters = pio.check_hhparams(args.hhparams)
else:
try:
hhparameters = pio.check_hhparams(pio.conf.HHBLITS_PARAMETERS)
except:
hhparameters = pio.check_hhparams('dmp')
# Define formats used
sources = pio.paths.sources()
n_sources = len(sources)
# Parse sequence and structure files
logger.info('Parsing sequence file...')
seq = cps.parseseqfile(inseq)
if len(seq) == 1:
for key in seq.keys():
pdbid = key.lower()
if len(seq[key].imer) == 1:
for key2 in seq[key].imer[key2]:
chid = key2
else:
raise Exception('More than one pdbid in sequence set.')
else:
raise Exception('More than one pdbid in sequence set.')
logger.info('Parsing structure file...')
structure = cps.parsestrfile(instr)[0][pdbid]
#logger.info('Parsing SIFTS database file...')
#sifts = cps.import_db(indb, pdb_in=pdbid)
# CROPPING AND RENUMBERING
if not skipexec[0]:
logger.info(
'Cropping and renumbering sequences, structures according to SIFTS database.'
)
psys.crops.runcrops(inseq, instr, indb, thuprot, dbuprot, outrootdir)
outpdbdir = os.path.join(outrootdir, pdbid, "")
pio.mdir(outpdbdir)
inseqc = os.path.join(outpdbdir, os.path.basename(inseq))
instrc = os.path.join(outpdbdir, os.path.basename(instr))
copyfile(inseq, inseqc)
copyfile(instr, instrc)
cmappath = os.path.join(os.path.splitext(inseqc), '.cropmap')
# MSA GENERATOR
cseqpath = os.path.join(outpdbdir, pdbid + '.crops.to_uniprot.fasta')
hhdir = os.path.join(outpdbdir, 'hhblits', '')
pio.mdir(hhdir)
neff = {}
if not skipexec[0] or not skipexec[1]:
if hhparameters == ['3', '0.001', 'inf', '50', '99']:
logger.info(
'Generating Multiple Sequence Alignment using DeepMetaPSICOV default parameters... [AS RECOMMENDED]'
)
elif hhparameters == ['2', '0.001', '1000', '0', '90']:
logger.info(
'Generating Multiple Sequence Alignment using HHBlits default parameters...'
)
else:
logger.info(
'Generating Multiple Sequence Alignment using user-custom parameters...'
)
if os.path.isfile(cmappath) and not skipexec[0]:
psys.msagen.runhhblits(cseqpath, hhparameters, hhdir)
cmsaa3mfile = os.path.splitext(
os.path.basename(cseqpath))[0] + ".msa.a3m"
cmsaa3mpath = os.path.join(hhdir, cmsaa3mfile)
neff['cropped'] = psys.msagen.msafilesgen(cmsaa3mpath)
neff['original'] = None
if not skipexec[1]:
logger.info(
' Repeating process for non-default sequence...')
neff['cropped'] = None
else:
logger.info(
' No cropped sequence found, using original sequence instead...'
)
if not os.path.isfile(cmappath) or not skipexec[1]:
psys.msagen.runhhblits(inseq, hhparameters, hhdir)
msaa3mfile = os.path.splitext(
os.path.basename(inseq))[0] + ".msa.a3m"
msaa3mpath = os.path.join(hhdir, msaa3mfile)
neff['original'] = psys.msagen.msafilesgen(msaa3mpath)
# DEEP META PSICOV RUN
if not skipexec[0] or not skipexec[1]:
logger.info(
'Generating contact prediction lists via DeepMetaPSICOV...')
dmpdir = os.path.join(outpdbdir, 'dmp', '')
pio.mdir(dmpdir)
if os.path.isfile(cmappath) and not skipexec[0]:
psys.dmp.rundmp(cseqpath, cmsaa3mpath, dmpdir)
if not skipexec[1]:
logger.info(
' Repeating process for non-default sequence...')
else:
logger.info(
' No cropped sequence found, using original sequence instead...'
)
if not os.path.isfile(cmappath) or not skipexec[1]:
psys.dmp.rundmp(inseq, msaa3mpath, dmpdir)
# INTERFACE GENERATION, PISA
cstrpath = os.path.join(outpdbdir, pdbid + '.oldids.crops.to_uniprot.pdb')
pisadir = os.path.join(outpdbdir, 'pisa', '')
n_ifaces = {}
if not skipexec[0] or not skipexec[1]:
logger.info('Generating interface files via PISA...')
if os.path.isfile(cmappath) and not skipexec[0]:
n_ifaces['cropped'] = psys.pisa.runpisa(cstrpath, pisadir)
if not skipexec[1]:
logger.info(
' Repeating process for non-default sequence...')
else:
n_ifaces['cropped'] = None
logger.info(
' No cropped sequence found, using original sequence instead...'
)
if not os.path.isfile(cmappath) or not skipexec[1]:
n_ifaces['original'] = psys.pisa.runpisa(instr, pisadir)
else:
n_ifaces['original'] = None
# CONTACT ANALYSIS AND MATCH
resultdir = os.path.join(outpdbdir, 'results', '')
logger.info('Opening output csv files...')
pdbcsvfile = os.path.join(resultdir, pdbid + ".evcovsignal.csv")
pio.outcsv.csvheader(pdbcsvfile)
if not csvexists:
pio.outcsv.csvheader(outcsvfile)
logger.info('Parsing contact predictions lists...')
ckseq = ckio.read(inseq, 'fasta')
conpred = {}
for source, attribs in sources.items():
for mode in ['cropped', 'original']:
seqfile = cseqpath if mode == 'cropped' else inseq
confile = (os.path.splitext(os.path.basename(seqfile))[0] +
attribs[1])
conpath = os.path.join(outpdbdir, attribs[0], confile)
cropmapping = cps.parsemapfile(cmappath)
if os.path.isfile(conpath):
conpred[mode][source] = ckio.read(conpath, attribs[2])[0]
for contact in conpred[mode][source]:
contact.res1_seq = cropmapping[pdbid][chid]['cropbackmap'][
contact.res1_seq]
contact.res2_seq = cropmapping[pdbid][chid]['cropbackmap'][
contact.res2_seq]
else:
conpred[mode][source] = None
# NOT SURE IT IS WORKING WITH SEVERAL CHAINS OF SAME SEQUENCE. CHECK EVERYTHING.
logger.info(' Parsing PISA interfaces...')
interfaces = {}
for mode in ['cropped', 'original']:
if n_ifaces[mode] is not None:
interfaces[mode] = []
for i in range(int(n_ifaces[mode])):
strfile = cstrpath if mode == 'cropped' else instr
pdbfilei = os.path.splitext(strfile)[0] + ".interface." + str(
i + 1) + ".pdb"
interfaces[mode].append(ckio.read(pdbfilei, 'pdb'))
else:
interfaces[mode] = None
# OUTPUT
# CODE TO PRINT NON-REPEATED LINES
# import csv
# rows = csv.reader(open("file.csv", "rb"))
# newrows = []
# for row in rows:
# if row not in newrows:
# newrows.append(row)
# writer = csv.writer(open("file.csv", "wb"))
# writer.writerows(newrows)
return
def main():
starttime=time.time()
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.crops_logger(level="info")
logger.info(pcl.welcome())
inseq=check_path(args.input_seqpath[0],'file')
indb=check_path(args.input_database[0],'file')
insprot=check_path(args.uniprot_threshold[1]) if args.uniprot_threshold is not None else None
minlen=float(args.uniprot_threshold[0]) if args.uniprot_threshold is not None else 0.0
targetlbl=ctg.target_format(indb,terms=args.terminals, th=minlen)
infixlbl=ctg.infix_gen(indb,terms=args.terminals)
if args.outdir is None:
outdir=check_path(os.path.dirname(inseq),'dir')
else:
outdir=check_path(os.path.join(args.outdir[0],''),'dir')
if args.sort is not None:
if (args.sort[0].lower()!='ncrops' and args.sort[0].lower()!='percent' and
args.sort[0].lower()!='ncropsin' and args.sort[0].lower()!='percentin'):
raise ValueError("Arguments for sorting option can only be either 'ncrops' or 'percent'.")
else:
sorter=args.sort[0].lower()
#############################################################
sources=["deepmetapsicov", "psicov"]
confiledir=["deepmetapsicov", "deepmetapsicov"]
confilesuffix=["deepmetapsicov.con","psicov"]
n_sources=len(sources)
# BEGIN: DEBUGGING ONLY
# Skip the execution of PISA, HHBLITS and DeepMetaPSICOV (?) [bool]
# SKIP_EXEC=[True, True, True]
SKIP_EXEC=[False, False, False]
# Skip the creation of new dir (?) [bool]
# SKIP_MKDIR=True
SKIP_MKDIR=False
# END: DEBUGGING ONLY
# Create output directory
if not SKIP_MKDIR:
pmo.mkdirout()
pmo.printout('Output directory created',extraline=True)
starttime=pmi2.gettime()
pmo.printout('Starting Time: '+ starttime[1].strftime("%-d %B %Y, %X") + ' UTC\n\n')
# Check that input values are correct
hhparameters=pmi1.hhparam('logit')
dumvar=pmi1.minneigh('logit')
dumvar=pmi1.scorethreshold("deepmetapsicov",'logit')
dumvar=pmi1.scorethreshold("psicov",'logit')
#####################################################################
##### MSA generator #################################################
pmo.printout('*********************************************************')
pmo.printout('*** MULTIPLE SEQUENCE ALIGNMENT *************************',extraline=True)
pmo.printout('Importing RCSB PDB sequence from fasta file...')
fasta_seq=conkit.io.read(pmin.SEQUENCE_PATH,"fasta")[0]
pmo.printout(fasta_seq)
if pmin.USE_BIOCHAIN:
pmo.printout('Obtaining Biological sequence from fasta file...')
if not SKIP_EXEC[1]:
biological_seq, seqpath, bio = pms.crop_fasta(fasta_seq,pmo.output_tmpdir("deepmetapsicov"))
if not bio:
shutil.copyfile(pmin.SEQUENCE_PATH, seqpath)
else:
tmpfile = pmi2.pdbid() + '.bio.fasta'
seqpath = os.path.join(pmo.output_tmpdir("deepmetapsicov"), tmpfile)
try:
biological_seq=conkit.io.read(seqpath,"fasta")[0]
bio=True
except:
bio=False
tmpfile = pmi2.pdbid() + '.fasta'
seqpath = os.path.join(pmo.output_tmpdir("deepmetapsicov"), tmpfile)
biological_seq=conkit.io.read(seqpath,"fasta")[0]
seq=biological_seq
tmpfile = pmi2.pdbid() + pms.biofile(bio)+ '.fasta'
newseqpath = os.path.join(pmo.output_dir(), tmpfile)
shutil.copyfile(seqpath, newseqpath)
if bio:
pmo.printout('Biological sequence:')
pmo.printout(biological_seq)
else:
pmo.printout('WARNING: Biological Sequence not available. Using RCSB PDB fasta sequence for analysis (including cloning artifacts)',errorlog=True, extraline=True) # LOGGING
else:
bio=False
pmo.printout('Using RCSB PDB fasta sequence for analysis (including cloning artifacts)', extraline=True)
seqpath = os.path.join(pmo.output_dir(), os.path.splitext(os.path.basename(pmin.SEQUENCE_PATH))[0])+ os.path.splitext(os.path.basename(pmin.SEQUENCE_PATH))[1]
shutil.copyfile(pmin.SEQUENCE_PATH, seqpath)
seqpath = os.path.join(pmo.output_tmpdir("deepmetapsicov"), os.path.splitext(os.path.basename(pmin.SEQUENCE_PATH))[0])+ os.path.splitext(os.path.basename(pmin.SEQUENCE_PATH))[1]
shutil.copyfile(pmin.SEQUENCE_PATH, seqpath)
seq=fasta_seq
pmo.printout('')
if pmin.HHBLITS_VIA_DMP:
pmo.printout("Using DeepMetaPSICOV's execution of HHblits...")
msapath=''
else:
if not SKIP_EXEC[1]:
pmo.printout('Creating Multiple Sequence Alignment with HHblits...')
msa, msapath = pmp.runhhblits(bio,param=hhparameters,spath=seqpath)
else:
pmo.printout('Reading Multiple Sequence Alignment...')
msafile = pmi2.pdbid()+pms.biofile(bio)+".msa.aln"
msapath = os.path.join(pmo.output_dir(), msafile)
msa = conkit.io.read(msapath,'jones')
msacovpath=os.path.splitext(msapath)[0]+".coverage.png"
msaformat='jones'
sit=0.7
#####################################################################
##### DeepMetaPSICOV: Contact Prediction ############################
pmo.printout('*********************************************************')
pmo.printout('*** CONTACT PREDICTION **********************************',extraline=True)
if not SKIP_EXEC[2]:
pmo.printout('Running DeepMetaPSICOV for Contact prediction list...')
else:
pmo.printout('Skipping DeepMetaPSICOV execution.')
if pmin.HHBLITS_VIA_DMP:
msa, msapath = pmp.rundmp(seqpath, msapath,skiphhblits=SKIP_EXEC[1],skipdmp=SKIP_EXEC[2]) # Input Sequence fasta file path, MSA file path
else:
if not SKIP_EXEC[2]:
pmp.rundmp(seqpath, msapath) # Input Sequence fasta file path, MSA file path
#####################################################################
##### PISA : OBTAIN INTERFACE PDB FILES #############################
pmo.printout('*********************************************************')
pmo.printout('*** INTERFACE IDENTIFICATION ****************************',extraline=True)
# Obtain Number of interfaces in PDB and produce Interface PDB files (including renumbered)
if not SKIP_EXEC[0]:
pmo.printout('Running PISA...')
n_interfaces = pmp.runpisa(bio)
for nif in range (n_interfaces):
pdbintpath = os.path.join(pmo.output_tmpdir("pisa"), pmi2.pdbid()+".interface."+str(nif+1)+".pdb")
pms.renumberpdbs(pdbintpath, fasta_seq, bio)
else:
pmo.printout('Reading PISA xml file...')
n_interfaces = pmp.n_int_xml()
#####################################################################
##### Interface Confidence Scores ###################################
pmo.printout('*********************************************************')
pmo.printout('*** SCORING INTERFACES **********************************',extraline=True)
pmo.printout('Importing contact prediction file ...',extraline=True)
conpred=[]
conpred_id=[]
conpredpath=[]
for n in range(n_sources):
conpredfile=pmi2.pdbid()+pms.biofile(bio)+"."+confilesuffix[n]
conpredpath.append(os.path.join(pmo.output_tmpdir(confiledir[n]), conpredfile))
conpred.append(conkit.io.read(conpredpath[n], 'psicov')[0])
conpred[n].sequence=seq
conpred[n].set_sequence_register()
conpred_id.append(conpred[n].id)
pmo.printout('Remove excluded contact scores ...',extraline=True)
for n in range(n_sources):
conpred[n]=pmc.filter_contacts(conpred[n],sources[n])
cntint=0
cntlig=0
cntunk=0
interfacetype=[0 for i in range(n_interfaces+1)]
scores = [[[0 for k in range(4)] for j in range(n_sources)] for i in range(n_interfaces)]
n_contacts_all = [[[0 for k in range(4)] for j in range(n_sources)] for i in range(n_interfaces)]
ndec=6
for nif in range(n_interfaces):
pmo.printout('Importing Interface '+str(nif+1)+' PDB file...')
maps=pmc.import_interfacepdb(nif,bio)
conpredInt=[]
pmo.printout(' Number of maps : ' + str(len(maps)) )
if (len(maps)==4):
for n in range(n_sources):
interfacetype[nif]="MM"
conpredInt.append(conpred[n].deepcopy())
if pmin.REMOVE_INTRA_CONTACTS:
pmo.printout(" Removing those contacts from interface that also appear as intramolecular contacts ...")
for element in [0,3]:
conpredInt[n]=pmc.remove_intra_contacts(conpredInt[n],maps[element])
n_contacts_all[nif][n][0]=conpredInt[n].ncontacts # Total number of contacts predicted for interface
n_contacts_all[nif][n][1]=maps[1].ncontacts # Total number of contacts from interface PDB file
#pmo.printout(conpredInt)
pmo.printout(maps[1])
cntint += 1
# Write contact list
pmo.printout(' Writing contact lists ...' )
if pmin.REMOVE_INTRA_CONTACTS:
for element in [0,3]:
chainid = str(1) if element == 0 else str(2)
confile=pmi2.pdbid()+".pdb.interface."+str(nif+1)+".intrachain"+chainid+".conkit.con"
tmppath = os.path.join(pmo.output_tmpdir("pisacov"), confile)
conkit.io.write(tmppath, 'psicov', hierarchy=maps[element])
confile=pmi2.pdbid()+".pdb.interface."+str(nif+1)+".conkit.con"
tmppath = os.path.join(pmo.output_tmpdir("pisacov"), confile)
conkit.io.write(tmppath, 'psicov', hierarchy=maps[1])
pmo.printout(' Matching contact prediction and interface contact lists ...',extraline=True )
contactpath=os.path.join(pmo.output_tmpdir("pisacov"), confile)
for n in range(n_sources):
conpredInt_pdb=pmc.pred_pdb_matchlist(conpredInt[n],contactpath)
conpredInt_pdb.id=" Matching contacts of " + conpred_id[n]
conpredInt_pdb.sequence = seq.deepcopy()
conpredInt_pdb.set_sequence_register()
if conpredInt_pdb.ncontacts == 0:
pmo.printout(" WARNING: No contacts found in the filtered interface "+str(nif)+" pdb file. SKIP this result type.",errorlog=True)
for sc in range(4):
scores[nif][n][sc]="***"
n_contacts_all[nif][n][3] = "***"
else:
map_matched, n_contacts_all[nif][n][2], scores[nif][n][0], scores[nif][n][1], scores[nif][n][2], scores[nif][n][3] = pmc.match_maps(conpredInt_pdb, maps[1])
for sc in range(4):
scores[nif][n][sc]=round(scores[nif][n][sc],ndec)
n_contacts_all[nif][n][3] = n_contacts_all[nif][n][2] / n_contacts_all[nif][n][1]
## Plot matched map
pngfile=pmi2.pdbid()+".interface."+str(nif+1)+"."+sources[n]+".con.png"
confile=pmi2.pdbid()+"."+sources[n]+".interface."+str(nif+1)+".conkit.con"
pngpath = os.path.join(pmo.output_dir(), pngfile)
fig = conkit.plot.ContactMapFigure(map_matched, reference=maps[1])
fig.savefig(pngpath, overwrite=True)
tmppath = os.path.join(pmo.output_dir(), confile)
conkit.io.write(tmppath, 'psicov',hierarchy=conpredInt_pdb)
plt.close('all')
elif (len(maps)==2):
interfacetype[nif]="Unk"
pmo.printout(" WARNING: Unexpected number of maps (2)", errorlog=True)
pmo.printout('----> REVISE CONTACT MAPS FOR THIS INTERFACE', errorlog=True,extraline=True)
cntunk += 1
cnt=0
for score in range(4):
for n in range(n_sources):
scores[nif][n][score]="***"
for mapn in maps:
confile=pmi2.pdbid()+".pdb."+str(cnt+1)+".interface."+str(nif+1)+".conkit.con" #### CHECK FILES CREATED WHEN FIXING THIS SECTION.
tmppath = os.path.join(pmo.output_dir(), confile)
conkit.io.write(tmppath, 'psicov', hierarchy=maps[1])
cnt += 1
elif (len(maps)==1):
interfacetype[nif]="LM"
pmo.printout(" ... Ligand-monomer interface detected")
pmo.printout(' SKIP', extraline=True)
cntlig += 1
for score in range(4):
for n in range(n_sources):
scores[nif][n][score]="***"
else:
interfacetype[nif]="Unk"
pmo.printout(" WARNING: Unexpected number of maps", errorlog=True)
pmo.printout(' SKIP', extraline=True, errorlog=True)
cntunk += 1
for score in range(4):
for n in range(n_sources):
scores[nif][n][score]="***"
#####################################################################
##### OUTPUT ########################################################
pmo.printout('*********************************************************')
pmo.printout('*** FINAL OUTPUT ****************************************',extraline=True)
pmo.printout('Printing final data to file...', extraline=True)
datfile=pmi2.pdbid()+".pisacov.out.dat"
datpath = os.path.join(pmo.output_dir(), datfile)
now = pmi2.gettime()
space=str(ndec+4)
with open(datpath, 'w') as out:
if bio:
out.write( "# PDB id: "+ pmi2.pdbid() + ' - Contact def: 0<d<8 Angstroms - Cloning artifacts removed from fasta file sequence - ' )
else:
out.write( "# PDB id: "+ pmi2.pdbid() + ' - Contact def: 0<d<8 Angstroms - Original fasta file sequence used - ' )
out.write( now[1].strftime("%-d %B %Y, %X") + ' UTC')
out.write('\n')
datahead='# IF_id IF_type n_pdb '
for n in range(n_sources):
datahead += 'n_'+sources[n]+' '
datahead += 'n_'+sources[n]+'/n_pdb '
datahead += 'Av_score_'+sources[n]+' '
datahead += 'Acc_score_'+sources[n]+' '
datahead += 'P_'+sources[n]+' '
out.write(datahead+'\n')
for nif in range(n_interfaces):
outformat='{:>4s}{:>4s}{:>4s}'
outstring = str(outformat.format(str(nif+1),interfacetype[nif],str(n_contacts_all[nif][0][1])))
for n in range(n_sources):
#outformat='{:>4s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}'
outformat='{:>4s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}{:>'+space+'s}'
outstring += str(outformat.format(str(n_contacts_all[nif][n][2]),str(scores[nif][n][0]),str(scores[nif][n][1]),str(scores[nif][n][2]),str(scores[nif][n][3])))
out.write(outstring+'\n')
sumlogfile=pmi2.pdbid()+".pisacov.out.summary.log"
sumlogpath = os.path.join(pmo.output_dir(), sumlogfile)
now = pmi2.gettime()
with open(sumlogpath, 'w') as out:
out.write('*********************************************************\n')
out.write('*** S U M M A R Y ** P I S A C O V ********************\n')
out.write('*********************************************************\n\n')
out.write(now[1].strftime("%-d %B %Y, %X") + ' UTC\n\n')
out.write('Protein PDB ID: '+pmi2.pdbid()+'\n\n')
out.write('*********************************************************\n')
out.write('--- Multiple Sequence Alignment ---\n\n')
outformat='{:<55s}{:<80s}'
out.write(outformat.format(' MSA file: ', msapath))
out.write('\n')
out.write(outformat.format(' MSA format: ', msaformat))
out.write('\n')
out.write(outformat.format(' Sequence Identity Threshold: ', str(sit) ))
out.write('\n')
out.write(outformat.format(' Length of the Target Sequence: ',str(msa.top_sequence.seq_len)))
out.write('\n')
out.write(outformat.format(' Total number of sequences: ', str(msa.nseq)))
out.write('\n')
out.write(outformat.format(' Number of Effective Sequences: ', str(msa.meff)))
out.write('\n')
out.write(outformat.format(' Proportion of Effective Sequences: ', str(round(100*msa.meff/msa.nseq,2))+' %'))
out.write('\n')
out.write(outformat.format(' Sequence Coverage Plot: ', msacovpath))
out.write('\n\n')
out.write(outformat.format(' MSA created with: ',pmin.HHSUITE_PATH))
out.write('\n')
out.write(outformat.format(' Reference database name: ', pmin.HHBLITS_DATABASE_NAME))
out.write('\n')
out.write(outformat.format(' Reference database path: ', pmin.HHBLITS_DATABASE_DIR))
out.write('\n')
out.write(outformat.format(' Number of iterations: ', str(pmi1.hhparam()[0])))
out.write('\n')
out.write(outformat.format(' E-value cutoff for inclusion in result alignment: ', str(pmi1.hhparam()[1])))
out.write('\n')
out.write(outformat.format(' Non-redundant sequences to keep: ', str(pmi1.hhparam()[2])))
out.write('\n')
out.write(outformat.format(' Minimum coverage with master sequence (%): ', str(pmi1.hhparam()[3])))
out.write('\n')
out.write(outformat.format(' Maximum pairwise sequence identity: ', str(pmi1.hhparam()[4])))
out.write('\n\n')
out.write('*********************************************************\n')
out.write('--- Contact Prediction ---\n\n')
outformat='{:<55s}{:<80s}'
out.write(outformat.format(' Contact prediction list(s) created with: ', pmin.DMP_PATH))
out.write('\n')
for n in range(n_sources):
out.write(outformat.format(' Contact prediction file ('+sources[n]+'): ', conpredpath[n]))
out.write('\n')
out.write(outformat.format(' Minimum distance within sequence (neigh. cutoff): ', str(pmi1.minneigh()) ))
out.write('\n')
for n in range(n_sources):
out.write(outformat.format(' Contact prediction score threshold ('+sources[n]+'): ', str(pmi1.scorethreshold(sources[n]))))
out.write('\n')
out.write('\n')
out.write('*********************************************************\n')
out.write('--- Interfaces ---\n\n')
outformat='{:<55s}{:<80s}'
out.write(outformat.format(' Total Number of Interfaces: ', str(n_interfaces) ))
out.write('\n')
out.write(outformat.format(' Number of Intramolecular interfaces: ', str(cntint) ))
out.write('\n')
out.write(outformat.format(' Number of Ligand-monomer interfaces: ', str(cntlig) ))
out.write('\n')
out.write(outformat.format(' Number of Unidentified interfaces: ', str(cntunk) ))
out.write('\n\n')
for nif in range(n_interfaces):
nif1=nif+1
out.write(' --- Interface '+str(nif1)+' ---\n')
if (interfacetype[nif] == "MM"):
out.write(outformat.format(' Interface type:','monomer - monomer (MM)'))
out.write('\n')
out.write(outformat.format(' Total number of intermolecular contacts (pdb): ', str(n_contacts_all[nif][0][1])))
out.write('\n')
for n in range(n_sources):
extral='\n\n' if n==n_sources-1 else '\n'
out.write(' + '+sources[n]+' Scores +\n')
out.write(outformat.format(' Number of True Positives : ', str(n_contacts_all[nif][n][2]) ))
out.write('\n')
out.write(outformat.format(' Proportion of True positives: ', str(n_contacts_all[nif][n][3])))
out.write('\n')
out.write(outformat.format(' Jaccard Index (|PDB ⋂ pred| / |PDB U pred|): ', str(scores[nif][n][0])))
out.write('\n')
out.write(outformat.format(' Average value of the scores of True Positives: ', str(scores[nif][n][1])))
out.write('\n')
out.write(outformat.format(' Sum of all the scores of True Positives: ', str(scores[nif][n][2])))
out.write('\n')
out.write(outformat.format(' Probabilistic score for whole interface: ', str(scores[nif][n][3])))
out.write(extral)
elif (interfacetype[nif] == "LM"):
out.write(outformat.format(' Interface type:','ligand - monomer (LM)'))
out.write('\n\n')
elif (interfacetype[nif] == "Unk"):
out.write(outformat.format(' Interface type:','unidentified (Unk)'))
out.write('\n\n')
pmo.printout('*********************************************************')
pmo.printout('*** S U M M A R Y ***************************************')
pmo.printout('*********************************************************', extraline=True)
pmo.printout('Protein PDB ID: %s' % pmi2.pdbid(), extraline=True)
pmo.printout('--- Multiple Sequence Alignment ---')
pmo.printout(' MSA file: ' + msapath)
pmo.printout(' MSA format: ' + msaformat)
pmo.printout(' Sequence Identity Threshold: '+ str (sit))
pmo.printout(' Length of the Target Sequence: ' + str(msa.top_sequence.seq_len))
pmo.printout(' Total number of sequences: ' + str(msa.nseq))
pmo.printout(' Number of Effective Sequences: ' + str(msa.meff))
pmo.printout(' Proportion of Effective Sequences: ' + str( round(100*msa.meff/msa.nseq,2))+' %')
pmo.printout(' Sequence Coverage Plot: ' + msacovpath, extraline=True)
pmo.printout(' MSA created with: ' + pmin.HHSUITE_PATH)
pmo.printout(' Reference database name: ' + pmin.HHBLITS_DATABASE_NAME)
pmo.printout(' Reference database path: ' + pmin.HHBLITS_DATABASE_DIR)
pmo.printout(' Number of iterations: ' + str(pmi1.hhparam()[0]))
pmo.printout(' E-value cutoff for inclusion in result alignment: ' + str(pmi1.hhparam()[1]))
pmo.printout(' Non-redundant sequences to keep: ' + str(pmi1.hhparam()[2]))
pmo.printout(' Minimum coverage with master sequence (%): ' + str(pmi1.hhparam()[3]))
pmo.printout(' Maximum pairwise sequence identity: ' + str(pmi1.hhparam()[4]), extraline=True)
pmo.printout('--- Contact Prediction ---')
pmo.printout(' Contact prediction list created with: ' + pmin.DMP_PATH)
for n in range(n_sources):
pmo.printout(' Contact prediction file ('+sources[n]+'): ' + conpredpath[n])
pmo.printout(' Minimum distance within sequence (neigh. cutoff): '+ str(pmi1.minneigh()))
for n in range(n_sources):
extral=True if n==n_sources-1 else False
pmo.printout(' Contact prediction score threshold ('+sources[n]+'): ' + str(pmi1.scorethreshold(sources[n])),extraline=extral)
pmo.printout('--- Interfaces ---')
pmo.printout(' Number of Interfaces: ' + str(n_interfaces) + ", of which:")
pmo.printout(' Number of Intramolecular interfaces: ' + str(cntint))
pmo.printout(' Number of Ligand-monomer interfaces: ' + str(cntlig))
pmo.printout(' Number of Unidentified interfaces: ' + str(cntunk),extraline=True)
for nif in range(n_interfaces):
nif1=nif+1
pmo.printout(' --- Interface '+str(nif1)+' ---',extraline=True)
if (interfacetype[nif] == "MM"):
pmo.printout(' Interface type: monomer - monomer (MM)')
pmo.printout(' Total number of intermolecular contacts (pdb): ' + str(n_contacts_all[nif][0][1]))
for n in range(n_sources):
extral=True if n==n_sources-1 else False
pmo.printout(' + Scores ('+sources[n]+'):')
pmo.printout(' Number of True Positives : ' + str(n_contacts_all[nif][n][2]))
pmo.printout(' Proportion of True positives: ' + str(n_contacts_all[nif][n][3]) )
pmo.printout(' Jaccard Index (|PDB ⋂ pred| / |PDB U pred|): ' + str(scores[nif][n][0]) )
pmo.printout(' Average value of the scores of True Positives: ' + str(scores[nif][n][1] ))
pmo.printout(' Sum of all the scores of True Positives: ' + str(scores[nif][n][2] ))
pmo.printout(' Probabilistic score for whole interface: ' + str(scores[nif][n][3] ),extraline=extral)
elif (interfacetype[nif] == "LM"):
pmo.printout(' Interface type: ligand - monomer (LM)',extraline=True)
elif (interfacetype[nif] == "Unk"):
pmo.printout(' Interface type: unidentified (Unk)',extraline=True)
#####################################################################
##### TIMEIT ########################################################
pmo.printout('*********************************************************')
pmo.printout('*** T I M I N G *****************************************')
pmo.printout('*********************************************************', extraline=True)
with open(sumlogpath, 'a') as out:
out.write('*********************************************************\n')
out.write('*** T I M I N G *****************************************\n')
out.write('*********************************************************\n\n')
endtime=pmi2.gettime()
totaltime=pmi2.readabletime(endtime[0]-starttime[0])
out.write('Starting Time: '+ starttime[1].strftime("%-d %B %Y, %X") + ' UTC\n\n')
out.write('Ending Time: '+ endtime[1].strftime("%-d %B %Y, %X") + ' UTC\n\n')
out.write('Process Wallclock time: '+str(endtime[0]-starttime[0])+' s\n')
out.write(' or, equivalently: ' + totaltime+'\n\n')
pmo.printout('Starting Time: '+ starttime[1].strftime("%-d %B %Y, %X")+' UTC', extraline=True)
pmo.printout('Ending Time: '+ endtime[1].strftime("%-d %B %Y, %X")+' UTC', extraline=True)
pmo.printout('Process Wallclock time: '+str(endtime[0]-starttime[0])+' s')
pmo.printout(' or, equivalently: ' + totaltime, extraline=True)
def main():
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
welcomemsg, starttime = pcl.welcome(command=__script__)
logger.info(welcomemsg)
kwlist = pco._default_keys()
configin = {}
if args.update_sifts_database is not None:
outpath = ppaths.check_path(args.update_sifts_database[0], 'either')
if os.path.isdir(outpath) is True:
outpath = os.path.join(outpath,
('pdb_chain_uniprot' + os.extsep + 'csv'))
pol.getsifts(outpath)
configin['SIFTS_PATH'] = outpath
else:
pass
if args.view_configuration is True:
print('** ' + __prog__ + " v." + __version__ +
' configuration file **' + os.linesep)
with open(pconf.__file__) as f:
for line in f:
line = _lineformat(line)
print(line, end='')
print('** End of configuration file **' + os.linesep)
return
else:
pass
if args.get_confpath is True:
print(pconf.__file__)
return
else:
pass
if args.conf_file is False:
newconf = pco._parse_conf()
else:
newconf = {}
if args.reset_hhblits_arguments is True:
newconf['HHBLITS_PARAMETERS'] = pco._default_values(
'HHBLITS_PARAMETERS')
else:
pass
if args.sifts_path is not None:
configin['SIFTS_PATH'] = args.sifts_path[0]
if args.pisa_path is not None:
configin['PISA_PATH'] = args.pisa_path[0]
if args.hhblits_path is not None:
configin['HHBLITS_PATH'] = args.hhblits_path[0]
if args.dmp_path is not None:
configin['DMP_PATH'] = args.dmp_path[0]
if args.uniclust_path is not None:
configin['UNICLUST_FASTA_PATH'] = args.uniclust_path[0]
if args.hhblits_arguments is not None:
configin['HHBLITS_PARAMETERS'] = args.hhblits_arguments[0]
if args.neighbours is not None:
configin['NEIGHBOURS_MINDISTANCE'] = args.neighbours[0]
configin['REMOVE_INTRA_CONTACTS'] = False
if isinstance(args.hhblits_location, list) is True:
configin['HHBLITS_DATABASE_NAME'] = args.hhblits_location[0]
configin['HHBLITS_DATABASE_DIR'] = args.hhblits_location[1]
compmsg = {
'SIFTS_PATH':
"Please, enter the path to the SIFTS csv file:\n",
'PISA_PATH':
"Please, enter the path to the PISA executable:\n",
'HHBLITS_PATH':
"Please, enter the path to the HHBLITS executable:\n",
'HHBLITS_DATABASE_NAME':
("Please, enter the name of the HHBLITS database name\n" +
"as requested by DeepMetaPSICOV (e.g. uniclust30_2018_08):\n"),
'HHBLITS_DATABASE_DIR':
"Please, enter the directory of the HHBLITS database:\n",
'DMP_PATH':
"Please, enter the path to the DeepMetaPSICOV executable:\n",
'HHBLITS_PARAMETERS':
("Please, enter the 5 HHBLITS parameters.\n" +
"#iterations, E-value cutoff, Non-redundant seqs to keep, " +
" MinimumCoverageWithMasterSeq(%), and MaxPairwiseSequenceIdentity.\n"
+ "Leave empty for DMP default (3, 0.001, 'inf', 50, 99).\n"),
'UNICLUST_FASTA_PATH':
("Please, enter the path to the UniClust fasta file.\n"
"An empty input will deactivate this option.\n"),
'NEIGHBOURS_MINDISTANCE':
("Press ENTER for intramolecular contacts " +
"to be removed from intermolecular contact lists.\n" +
"Otherwise, enter the minimum distance to be cosidered " +
"between neighbours. This will override the removal of" +
"intramolecular contacts that will be now ignored.\n")
}
for keystr in kwlist[:-1]:
if keystr in configin or args.conf_file is True:
if args.conf_file is True:
while True:
newval = input(compmsg[keystr])
try:
newconf[keystr] = pco._check_input(newval, keystr)
if keystr == 'NEIGHBOURS_MINDISTANCE':
if newval is None or newval == "":
newconf['REMOVE_INTRA_CONTACTS'] = True
else:
newconf['REMOVE_INTRA_CONTACTS'] = False
except Exception:
print("Not a valid input. Please, try again:")
else:
break
else:
newconf[keystr] = pco._check_input(configin[keystr], keystr)
if keystr == 'NEIGHBOURS_MINDISTANCE':
newconf['REMOVE_INTRA_CONTACTS'] = pco._check_input(
configin['REMOVE_INTRA_CONTACTS'],
'REMOVE_INTRA_CONTACTS')
else:
pass
if 'REMOVE_INTRA_CONTACTS' in configin:
newconf['REMOVE_INTRA_CONTACTS'] = pco._check_input(
configin['REMOVE_INTRA_CONTACTS'], 'REMOVE_INTRA_CONTACTS')
if newconf['REMOVE_INTRA_CONTACTS'] is True:
newconf['NEIGHBOURS_MINDISTANCE'] == 2
else:
pass
else:
pass
_outconffile(newconf)
endmsg = pcl.ok(starttime, command=__script__)
logger.info(endmsg)
return
def main():
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
welcomemsg, starttime = pcl.welcome(command=__script__)
logger.info(welcomemsg)
# PARSE CONFIGURATION FILE:
invals = pco._initialise_inputs()
invals['INSEQ'] = None
invals['INSTR'] = None
invals['ALTDB'] = None
invals['OUTROOT'] = None
invals['OUTCSVPATH'] = None
invals['UPTHRESHOLD'] = None
# READ INPUT ARGUMENTS
invals['INSEQ'] = ppaths.check_path(args.seqpath[0], 'file')
invals['INSTR'] = ppaths.check_path(args.crystalpath[0], 'file')
if args.hhblits_arguments is not None:
invals['HHBLITS_PARAMETERS'] = pco._check_hhparams(
args.hhblits_arguments)
else:
pass
if args.uniprot_threshold is not None:
try:
invals['UPTHRESHOLD'] = float(args.uniprot_threshold[0])
except ValueError:
logger.critical('Uniprot threshold given not valid.')
if invals['UNICLUST_FASTA_PATH'] is None:
invals['UNICLUST_FASTA_PATH'] = pco._uniurl
else:
pass
if args.skip_conpred is True:
skipexec = True
if (args.hhblits_arguments is not None
or args.uniprot_threshold is not None):
logger.info(
'HHblits, UniProt threshold parameters given bypassed by --skip_conpred'
)
else:
skipexec = False
cropping = args.remove_insertions
scoring = [cropping, not cropping]
if args.outdir is None:
invals['OUTROOT'] = ppaths.check_path(os.path.dirname(invals['INSEQ']))
else:
invals['OUTROOT'] = ppaths.check_path(os.path.join(args.outdir[0], ''))
ppaths.mdir(invals['OUTROOT'])
invals['OUTCSVPATH'] = []
if args.collection_file is None:
invals['OUTCSVPATH'].append(
ppaths.check_path(
os.path.join(invals['OUTROOT'],
("evcovsignal" + os.extsep + "cropped" +
os.extsep + "pisacov" + os.extsep + "csv"))))
invals['OUTCSVPATH'].append(
ppaths.check_path(
os.path.join(invals['OUTROOT'],
("evcovsignal" + os.extsep + "full" + os.extsep +
"pisacov" + os.extsep + "csv"))))
else:
if cropping is True:
invals['OUTCSVPATH'].append(
ppaths.check_path(args.collection_file[0]))
invals['OUTCSVPATH'].append(
ppaths.check_path(
os.path.splitext(args.collection_file[0])[0] + os.extsep +
'full' + os.extsep +
os.path.splitext(args.collection_file[0])[1]))
else:
invals['OUTCSVPATH'].append(None)
invals['OUTCSVPATH'].append(
ppaths.check_path(args.collection_file[0]))
# Define formats used
sources = pco._sources()
# Parse sequence and structure files
logger.info('Parsing sequence file...')
seqs = cps.parseseqfile(invals['INSEQ'])
logger.info('Parsing structure file...')
strs, filestrs = cps.parsestrfile(invals['INSTR'])
if len(seqs) == 1 or len(strs) == 1:
if len(seqs) == 1:
for key in seqs:
pdbid = key
elif len(seqs) > 1 and len(strs) == 1:
for key in strs:
for key2 in seqs:
if key.upper() == key2.upper():
pdbid = key.upper()
else:
if key2.upper() in key.upper():
pdbid = key2.upper()
else:
raise Exception(
'More than one pdbid in sequence and/or structure set.')
seq = seqs[pdbid]
#structure = strs[pdbid]
# CROPPING AND RENUMBERING
outpdbdir = os.path.join(invals['OUTROOT'], pdbid, "")
instrc = os.path.join(invals['OUTROOT'], pdbid,
os.path.basename(invals['INSTR']))
fseq = {}
fmsa = {}
if skipexec is False:
if cropping is True:
logger.info('Cropping and renumbering sequences, ' +
'structures according to SIFTS database.')
logger.info(pcl.running('CROPS-cropstr'))
itime = datetime.datetime.now()
psc.runcrops(invals['INSEQ'], invals['INSTR'],
invals['SIFTS_PATH'], invals['UPTHRESHOLD'],
invals['UNICLUST_FASTA_PATH'], invals['OUTROOT'])
logger.info(pcl.running('CROPS-cropstr', done=itime))
else:
logger.info('Renumbering structure ' +
'according to position in sequence.')
logger.info(pcl.running('CROPS-renumber'))
itime = datetime.datetime.now()
psc.renumcrops(invals['INSEQ'], invals['INSTR'], invals['OUTROOT'])
logger.info(pcl.running('CROPS-renumber', done=itime))
ppaths.mdir(outpdbdir)
if cropping is False:
psc.splitseqs(invals['INSEQ'], outpdbdir)
copyfile(invals['INSTR'], instrc)
for i, iseq in seq.imer.items():
fiseq = pdbid + '_' + i + '.fasta'
fseq[i] = os.path.join(invals['OUTROOT'], pdbid, fiseq)
fiseq = pdbid + '_' + i + '.msa.aln'
fmsa[i] = os.path.join(invals['OUTROOT'], pdbid, 'hhblits', fiseq)
if skipexec is False:
iseq.dump(fseq[i])
# Parse cropped sequences and maps
if cropping is True:
amap = {}
fcropseq = {}
fcropmsa = {}
for i, iseq in seq.imer.items():
fprefix = pdbid + '_' + i + '.crops.to_uniprot'
fmap = os.path.join(invals['OUTROOT'], pdbid,
fprefix + os.extsep + 'cropmap')
amap.update(cps.parsemapfile(fmap)[pdbid])
fcropseq[i] = os.path.join(invals['OUTROOT'], pdbid,
fprefix + os.extsep + 'fasta')
fcropmsa[i] = os.path.join(
invals['OUTROOT'], pdbid, 'hhblits',
(fprefix + os.extsep + 'msa' + os.extsep + 'aln'))
seq.set_cropmaps(amap, cropmain=True)
# EXECUTION OF EXTERNAL PROGRAMS
hhdir = os.path.join(invals['OUTROOT'], pdbid, 'hhblits', '')
dmpdir = os.path.join(invals['OUTROOT'], pdbid, 'dmp', '')
pisadir = os.path.join(invals['OUTROOT'], pdbid, 'pisa', '')
fstr = os.path.join(
invals['OUTROOT'],
(pdbid + os.extsep + 'crops' + os.extsep + 'seq' + os.extsep + 'pdb'))
if cropping:
fcropstr = os.path.join(
invals['OUTROOT'], pdbid,
(pdbid + os.extsep + 'crops' + os.extsep + 'oldids' + os.extsep +
'to_uniprot' + os.path.splitext(invals['INSTR'])[1]))
if skipexec is False:
# MSA GENERATOR
ppaths.mdir(hhdir)
if invals['HHBLITS_PARAMETERS'] == ['3', '0.001', 'inf', '50', '99']:
logger.info(
'Generating Multiple Sequence Alignment using DeepMetaPSICOV default parameters... [AS RECOMMENDED]'
)
elif invals['HHBLITS_PARAMETERS'] == ['2', '0.001', '1000', '0', '90']:
logger.info(
'Generating Multiple Sequence Alignment using HHBlits default parameters...'
)
else:
logger.info(
'Generating Multiple Sequence Alignment using user-custom parameters...'
)
for i, iseq in seq.imer.items():
sfile = fcropseq[i] if cropping is True else fseq[i]
afile = fcropmsa[i] if cropping is True else fmsa[i]
logger.info(pcl.running('HHBlits'))
itime = datetime.datetime.now()
themsa = psm.runhhblits(sfile, invals['HHBLITS_PARAMETERS'], hhdir)
logger.info(pcl.running('HHBlits', done=itime))
if cropping is True:
iseq.cropmsa = themsa
if iseq.ncrops() == 0:
iseq.msa = iseq.cropmsa
logger.info(' Cropped sequence ' + iseq.oligomer_id +
'_' + iseq.name +
' is identical to original sequence.')
continue
else:
pass
else:
iseq.msa = themsa
# DEEP META PSICOV RUN
ppaths.mdir(dmpdir)
if skipexec is False:
logger.info(
'Generating contact prediction lists via DeepMetaPSICOV...')
for i, iseq in seq.imer.items():
sfile = fcropseq[i] if cropping is True else fseq[i]
afile = fcropmsa[i] if cropping is True else fmsa[i]
nsfile = os.path.join(dmpdir, os.path.basename(sfile))
if sfile != nsfile:
copyfile(sfile, nsfile)
logger.info(pcl.running('DeepMetaPSICOV'))
itime = datetime.datetime.now()
psd.rundmp(nsfile, afile, dmpdir)
logger.info(pcl.running('DeepMetaPSICOV', done=itime))
# INTERFACE GENERATION, PISA
ppaths.mdir(pisadir)
if skipexec is False:
logger.info('Generating interface files via PISA...')
sfile = fcropstr if cropping is True else fstr
logger.info(pcl.running('PISA'))
itime = datetime.datetime.now()
iflist = psp.runpisa(sfile, pisadir, sessionid=pdbid)
logger.info(pcl.running('PISA', done=itime))
endmsg = pcl.ok(starttime, command=__script__)
logger.info(endmsg)
return
def main():
starttime = time.time()
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
welcomemsg, starttime = pcl.welcome(command=__script__)
logger.info(welcomemsg)
# PARSE CONFIGURATION FILE:
invals = pco._initialise_inputs()
invals['INSEQ'] = None
invals['INIFS'] = None
invals['OUTROOT'] = None
invals['OUTCSVPATH'] = None
# READ INPUT ARGUMENTS
invals['INSEQ'] == ppaths.check_path(args.seqpath[0], 'file')
invals['INIFS'] = []
args.remove_insertions = False
for fp in args.dimers:
if '*' in fp:
invals['INIFS'] += ppaths.check_wildcard(fp)
else:
invals['INIFS'].append(ppaths.check_path(fp, 'file'))
invals['INIFS'] = list(dict.fromkeys(invals['INIFS']))
if args.hhblits_arguments is not None:
invals['HHBLITS_PARAMETERS'] = pco._check_hhparams(
args.hhblits_arguments)
else:
pass
if args.skip_conpred is True:
skipexec = True
if args.hhblits_arguments is not None:
logger.info('HHblits parameters given bypassed by --skip_conpred')
else:
skipexec = False
if args.outdir is None:
invals['OUTROOT'] = ppaths.check_path(os.path.dirname(invals['INSEQ']))
else:
invals['OUTROOT'] = ppaths.check_path(os.path.join(args.outdir[0], ''))
ppaths.mdir(invals['OUTROOT'])
if args.collection_file is None:
invals['OUTCSVPATH'] = ppaths.check_path(
os.path.join(invals['OUTROOT'],
("evcovsignal" + os.extsep + "full" + os.extsep +
"pisacov" + os.extsep + "csv")))
else:
invals['OUTCSVPATH'] = ppaths.check_path(args.collection_file[0])
if os.path.isfile(invals['OUTCSVPATH']) is False:
pic.csvheader(invals['OUTCSVPATH'], cropped=False)
# Define formats used
sources = pco._sources()
# Parse sequence and structure files
logger.info('Parsing sequence file...')
seqs = cps.parseseqfile(invals['INSEQ'])
if len(seqs) == 1:
if len(seqs) == 1:
for key in seqs:
pdbid = key.lower()
else:
raise Exception('More than one pdbid in sequence set.')
seq = seqs[pdbid]
outpdbdir = os.path.join(invals['OUTROOT'], pdbid, "")
# RENUMBERING
fseq = {}
fmsa = {}
if skipexec is False:
if invals['INIFS'] is not None:
logger.info('Renumbering interfaces provided ' +
'according to position in sequence.')
for path in invals['INIFS']:
instrc = os.path.join(invals['OUTROOT'], pdbid,
os.path.basename(path))
logger.info(pcl.running('CROPS-renumber'))
itime = datetime.datetime.now()
psc.renumcrops(invals['INSEQ'], path, invals['OUTROOT'])
copyfile(path, instrc)
logger.info(pcl.running('CROPS-renumber', done=itime))
ppaths.mdir(outpdbdir)
for i, iseq in seq.imer.items():
fiseq = pdbid + '_' + i + os.extsep + 'fasta'
fseq[i] = os.path.join(invals['OUTROOT'], pdbid, fiseq)
fiseq = pdbid + '_' + i + os.extsep + 'msa' + os.extsep + 'aln'
fmsa[i] = os.path.join(invals['OUTROOT'], pdbid, 'hhblits', fiseq)
if skipexec is False:
iseq.dump(fseq[i])
# EXECUTION OF EXTERNAL PROGRAMS
hhdir = os.path.join(invals['OUTROOT'], pdbid, 'hhblits', '')
dmpdir = os.path.join(invals['OUTROOT'], pdbid, 'dmp', '')
fstr = []
for file in invals['INIFS']:
fstr.append(
os.path.join(
invals['OUTROOT'],
(os.path.splitext(os.path.basename(file))[0] + os.extsep +
'crops' + os.extsep + 'seq' + os.extsep + 'pdb')))
if skipexec is False:
# MSA GENERATOR
ppaths.mdir(hhdir)
if invals['HHBLITS_PARAMETERS'] == ['3', '0.001', 'inf', '50', '99']:
logger.info(
'Generating Multiple Sequence Alignment using DeepMetaPSICOV default parameters... [AS RECOMMENDED]'
)
elif invals['HHBLITS_PARAMETERS'] == ['2', '0.001', '1000', '0', '90']:
logger.info(
'Generating Multiple Sequence Alignment using HHBlits default parameters...'
)
else:
logger.info(
'Generating Multiple Sequence Alignment using user-custom parameters...'
)
for i, iseq in seq.imer.items():
sfile = fseq[i]
afile = fmsa[i]
logger.info(pcl.running('HHBlits'))
itime = datetime.datetime.now()
themsa = psm.runhhblits(sfile, invals['HHBLITS_PARAMETERS'], hhdir)
logger.info(pcl.running('HHBlits', done=itime))
iseq.msa = themsa
# DEEP META PSICOV RUN
logger.info(
'Generating contact prediction lists via DeepMetaPSICOV...')
ppaths.mdir(dmpdir)
for i, iseq in seq.imer.items():
sfile = fseq[i]
afile = fmsa[i]
nsfile = os.path.join(dmpdir, os.path.basename(sfile))
if sfile != nsfile:
copyfile(sfile, nsfile)
logger.info(pcl.running('DeepMetaPSICOV'))
itime = datetime.datetime.now()
psd.rundmp(nsfile, afile, dmpdir)
logger.info(pcl.running('DeepMetaPSICOV', done=itime))
# GENERATE INTERFACE LIST
iflist = []
for filepath in fstr:
ifname = os.path.splitext(os.path.basename(filepath))[0]
iflist.append(pci.interface(name=ifname))
# CONTACT ANALYSIS AND MATCH
logger.info('Opening output csv files...')
resultdir = os.path.join(invals['OUTROOT'], pdbid, 'pisacov', '')
ppaths.mdir(resultdir)
csvfile = os.path.join(
resultdir, (pdbid + os.extsep + "evcovsignal" + os.extsep + "full" +
os.extsep + "pisacov" + os.extsep + "csv"))
pic.csvheader(csvfile, cropped=False, pisascore=False)
logger.info('Parsing sequence files...')
for i, fpath in fseq.items():
seq.imer[i].seqs['conkit'] = ckio.read(fpath, 'fasta')[0]
seq.imer[i].biotype = csq.guess_type(seq.imer[i].seqs['mainseq'])
logger.info('Parsing contact predictions lists...')
conpred = {}
matches = []
for s in seq.imer:
if s not in conpred:
conpred[s] = {}
for source, attribs in sources.items():
fc = os.path.splitext(os.path.basename(fseq[s]))[0]
fc += attribs[1]
confile = os.path.join(invals['OUTROOT'], pdbid, attribs[0], fc)
conpred[s][source] = ckio.read(confile, attribs[2])[0]
logger.info('Parsing crystal structure contacts...')
for i in range(len(iflist)):
inputmap = ckio.read(fstr[i], 'pdb')
if len(inputmap) == 4:
chnames = list(iflist[i].chains.keys())
chtypes = list(iflist[i].chains.values())
if (seq.whatseq(chnames[0]) != seq.whatseq(chnames[1])
or (chtypes[0] != 'Protein' or chtypes[1] != 'Protein')):
if chtypes[0] != "Protein" or chtypes[1] != "Protein":
logger.info(
'Interface ' + str(i) +
' is not a Protein-Protein interface. Ignoring.')
else:
logger.info('Interface ' + str(i) +
' is not a homodimer. Ignoring.')
iflist[i].structure = None
matches.append(None)
continue
s = seq.whatseq(chnames[0])
try:
iflist[i].structure = []
for m in range(len(inputmap)):
iflist[i].structure.append(inputmap[m].as_contactmap())
iflist[i].structure[m].id = inputmap[m].id
except Exception:
for m in range(len(inputmap)):
iflist[i].structure.append(inputmap[m]) # ConKit LEGACY.
matches.append({})
for source, attribs in sources.items():
matches[i][source] = pcc.contact_atlas(
name=pdbid + '_' + str(s),
conpredmap=conpred[s][source],
strmap=iflist[i].structure,
sequence=seq.imer[s],
removeintra=True)
else:
iflist[i].structure = None
matches.append(None)
continue
logger.info('Computing results and writing them to file...')
for i in range(len(iflist)):
if matches[i] is None:
continue
results = [pdbid, str(i + 1)]
results.append(matches[i]['psicov'].chain1)
results.append(matches[i]['psicov'].chain2)
sid = seq.whatseq(matches[i]['psicov'].chain1)
results.append(str(sid))
results.append(str(seq.imer[sid].length()))
results.append(str(seq.imer[sid].cropmsa.meff))
results.append(str(seq.imer[sid].ncrops()))
results.append(str(seq.imer[sid].full_length()))
for source, attribs in sources.items():
appresults = pcs.list_scores(matches[i][source], tag=source)
results += appresults
pic.lineout(results, csvfile)
pic.lineout(results, invals['OUTCSVPATH'])
endmsg = pcl.ok(starttime, command=__script__)
logger.info(endmsg)
return
def main():
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
welcomemsg, starttime = pcl.welcome(command=__script__)
logger.info(welcomemsg)
# PARSE CONFIGURATION FILE:
invals = pco._initialise_inputs()
invals['INSEQ'] = None
invals['INSTR'] = None
invals['ALTDB'] = None
invals['OUTROOT'] = None
invals['OUTCSVPATH'] = None
invals['UPTHRESHOLD'] = None
# READ INPUT ARGUMENTS
invals['INSEQ'] = ppaths.check_path(args.seqpath[0], 'file')
invals['INSTR'] = ppaths.check_path(args.crystalpath[0], 'file')
if args.hhblits_arguments is not None:
invals['HHBLITS_PARAMETERS'] = pco._check_hhparams(
args.hhblits_arguments)
else:
pass
if args.uniprot_threshold is not None:
try:
invals['UPTHRESHOLD'] = float(args.uniprot_threshold[0])
except ValueError:
logger.critical('Uniprot threshold given not valid.')
if invals['UNICLUST_FASTA_PATH'] is None:
invals['UNICLUST_FASTA_PATH'] = pco._uniurl
else:
pass
if args.skip_conpred is True:
skipexec = True
if (args.hhblits_arguments is not None
or args.uniprot_threshold is not None):
logger.info(
'HHblits, UniProt threshold parameters given bypassed by --skip_conpred'
)
else:
skipexec = False
cropping = args.remove_insertions
scoring = [cropping, not cropping]
if args.outdir is None:
invals['OUTROOT'] = ppaths.check_path(os.path.dirname(invals['INSEQ']))
else:
invals['OUTROOT'] = ppaths.check_path(os.path.join(args.outdir[0], ''))
ppaths.mdir(invals['OUTROOT'])
invals['OUTCSVPATH'] = []
if args.collection_file is None:
invals['OUTCSVPATH'].append(
ppaths.check_path(
os.path.join(invals['OUTROOT'],
("evcovsignal" + os.extsep + "cropped" +
os.extsep + "pisacov" + os.extsep + "csv"))))
invals['OUTCSVPATH'].append(
ppaths.check_path(
os.path.join(invals['OUTROOT'],
("evcovsignal" + os.extsep + "full" + os.extsep +
"pisacov" + os.extsep + "csv"))))
else:
if cropping is True:
invals['OUTCSVPATH'].append(
ppaths.check_path(args.collection_file[0]))
invals['OUTCSVPATH'].append(
ppaths.check_path(
os.path.splitext(args.collection_file[0])[0] + os.extsep +
'full' + os.extsep +
os.path.splitext(args.collection_file[0])[1]))
else:
invals['OUTCSVPATH'].append(None)
invals['OUTCSVPATH'].append(
ppaths.check_path(args.collection_file[0]))
if args.plot_formats is None:
plotformats = {'png'}
else:
plotformats = set()
for element in args.plot_formats:
if element.lower() in {'png', 'eps', 'dat'}:
plotformats.add(element.lower())
# Define formats used
sources = pco._sources()
# Parse sequence and structure files
logger.info('Parsing sequence file...')
# seqs = cps.parseseqfile(invals['INSEQ'])
seqs = pio.read(invals['INSEQ'], 'fasta')
logger.info('Parsing structure file...')
# strs, filestrs = cps.parsestrfile(invals['INSTR'])
strs, filestrs = pio.read(invals['INSTR'], 'pdb')
if len(seqs) == 1 or len(strs) == 1:
if len(seqs) == 1:
for key in seqs:
pdbid = key
elif len(seqs) > 1 and len(strs) == 1:
for key in strs:
for key2 in seqs:
if key.upper() == key2.upper():
pdbid = key.upper()
else:
if key2.upper() in key.upper():
pdbid = key2.upper()
else:
raise Exception(
'More than one pdbid in sequence and/or structure set.')
seq = seqs[pdbid]
#structure = strs[pdbid]
# CROPPING AND RENUMBERING
outpdbdir = os.path.join(invals['OUTROOT'], pdbid, "")
instrc = os.path.join(invals['OUTROOT'], pdbid,
os.path.basename(invals['INSTR']))
fseq = {}
fmsa = {}
if skipexec is False:
if cropping is True:
logger.info('Cropping and renumbering sequences, ' +
'structures according to SIFTS database.')
logger.info(pcl.running('CROPS-cropstr'))
itime = datetime.datetime.now()
psc.runcrops(invals['INSEQ'], invals['INSTR'],
invals['SIFTS_PATH'], invals['UPTHRESHOLD'],
invals['UNICLUST_FASTA_PATH'], invals['OUTROOT'])
logger.info(pcl.running('CROPS-cropstr', done=itime))
else:
logger.info('Renumbering structure ' +
'according to position in sequence.')
logger.info(pcl.running('CROPS-renumber'))
itime = datetime.datetime.now()
psc.renumcrops(invals['INSEQ'], invals['INSTR'], invals['OUTROOT'])
logger.info(pcl.running('CROPS-renumber', done=itime))
ppaths.mdir(outpdbdir)
copyfile(invals['INSTR'], instrc)
for i, iseq in seq.imer.items():
fiseq = pdbid + '_' + i + '.fasta'
fseq[i] = os.path.join(invals['OUTROOT'], pdbid, fiseq)
fiseq = pdbid + '_' + i + '.msa.aln'
fmsa[i] = os.path.join(invals['OUTROOT'], pdbid, 'hhblits', fiseq)
if skipexec is False:
iseq.dump(fseq[i])
# Parse cropped sequences and maps
if cropping is True:
amap = {}
fcropseq = {}
fcropmsa = {}
for i, iseq in seq.imer.items():
fprefix = pdbid + '_' + i + '.crops.to_uniprot'
fmap = os.path.join(invals['OUTROOT'], pdbid,
fprefix + os.extsep + 'cropmap')
amap.update(cps.parsemapfile(fmap)[pdbid])
fcropseq[i] = os.path.join(invals['OUTROOT'], pdbid,
fprefix + os.extsep + 'fasta')
fcropmsa[i] = os.path.join(
invals['OUTROOT'], pdbid, 'hhblits',
(fprefix + os.extsep + 'msa' + os.extsep + 'aln'))
seq.set_cropmaps(amap, cropmain=True)
if iseq.ncrops() == 0:
logger.info(' Cropped sequence ' + iseq.oligomer_id + '_' +
iseq.name +
' is identical to the original sequence.')
else:
logger.info(' Cropped sequence ' + iseq.oligomer_id + '_' +
iseq.name + ' is ' + str(iseq.ncrops()) +
' residues ' +
'shorter than the original sequence.')
# EXECUTION OF EXTERNAL PROGRAMS
hhdir = os.path.join(invals['OUTROOT'], pdbid, 'hhblits', '')
dmpdir = os.path.join(invals['OUTROOT'], pdbid, 'dmp', '')
pisadir = os.path.join(invals['OUTROOT'], pdbid, 'pisa', '')
fstr = os.path.join(
invals['OUTROOT'],
(pdbid + os.extsep + 'crops' + os.extsep + 'seq' + os.extsep + 'pdb'))
if cropping:
fcropstr = os.path.join(
invals['OUTROOT'], pdbid,
(pdbid + os.extsep + 'crops' + os.extsep + 'oldids' + os.extsep +
'to_uniprot' + os.path.splitext(invals['INSTR'])[1]))
if skipexec is False:
# MSA GENERATOR
ppaths.mdir(hhdir)
if invals['HHBLITS_PARAMETERS'] == ['3', '0.001', 'inf', '50', '99']:
logger.info(
'Generating Multiple Sequence Alignment using DeepMetaPSICOV default parameters... [AS RECOMMENDED]'
)
elif invals['HHBLITS_PARAMETERS'] == ['2', '0.001', '1000', '0', '90']:
logger.info(
'Generating Multiple Sequence Alignment using HHBlits default parameters...'
)
else:
logger.info(
'Generating Multiple Sequence Alignment using user-custom parameters...'
)
for i, iseq in seq.imer.items():
sfile = fcropseq[i] if cropping is True else fseq[i]
afile = fcropmsa[i] if cropping is True else fmsa[i]
logger.info(pcl.running('HHBlits'))
itime = datetime.datetime.now()
themsa = psm.runhhblits(sfile, invals['HHBLITS_PARAMETERS'], hhdir)
logger.info(pcl.running('HHBlits', done=itime))
if cropping is True:
iseq.cropmsa = themsa
if iseq.ncrops() == 0:
iseq.msa = iseq.cropmsa
continue
else:
pass
else:
iseq.msa = themsa
# DEEP META PSICOV RUN
ppaths.mdir(dmpdir)
if skipexec is False:
logger.info(
'Generating contact prediction lists via DeepMetaPSICOV...')
for i, iseq in seq.imer.items():
sfile = fcropseq[i] if cropping is True else fseq[i]
afile = fcropmsa[i] if cropping is True else fmsa[i]
nsfile = os.path.join(dmpdir, os.path.basename(sfile))
if sfile != nsfile:
copyfile(sfile, nsfile)
logger.info(pcl.running('DeepMetaPSICOV'))
itime = datetime.datetime.now()
psd.rundmp(nsfile, afile, dmpdir)
logger.info(pcl.running('DeepMetaPSICOV', done=itime))
# INTERFACE GENERATION, PISA
ppaths.mdir(pisadir)
if skipexec is False:
logger.info('Generating interface files via PISA...')
sfile = fcropstr if cropping is True else fstr
logger.info(pcl.running('PISA'))
itime = datetime.datetime.now()
iflist = psp.runpisa(sfile, pisadir, sessionid=pdbid)
logger.info(pcl.running('PISA', done=itime))
# READ DATA IF SKIPEXEC USED:
if skipexec is True:
logger.info('Parsing already generated files...')
for i, iseq in seq.imer.items():
sfile = fcropstr if cropping is True else fstr
afile = fcropmsa[i] if cropping is True else fmsa[i]
if cropping is True:
# iseq.cropmsa = ckio.read(afile, 'jones')
iseq.cropmsa = pio.read(afile, 'jones')
if iseq.ncrops() == 0:
scoring[1] = True
# iseq.msa = ckio.read(afile, 'jones')
iseq.msa = ckio.read(afile, 'jones')
else:
# iseq.msa = ckio.read(afile, 'jones')
iseq.msa = pio.read(afile, 'jones')
ixml = os.path.join(pisadir,
(os.path.splitext(os.path.basename(sfile))[0] +
os.extsep + 'interface' + os.extsep + 'xml'))
axml = os.path.join(pisadir,
(os.path.splitext(os.path.basename(sfile))[0] +
os.extsep + 'assembly' + os.extsep + 'xml'))
iflist = pci.parse_interface_xml(ixml, axml)
# CONTACT ANALYSIS AND MATCH
logger.info('Opening output csv files...')
resultdir = os.path.join(invals['OUTROOT'], pdbid, 'pisacov', '')
ppaths.mdir(resultdir)
csvfile = []
csvfile.append(
os.path.join(resultdir,
(pdbid + os.extsep + "evcovsignal" + os.extsep +
"cropped" + os.extsep + "pisacov" + os.extsep + "csv")))
csvfile.append(
os.path.join(resultdir,
(pdbid + os.extsep + "evcovsignal" + os.extsep + "full" +
os.extsep + "pisacov" + os.extsep + "csv")))
for n in range(2):
if scoring[n] is True:
cpd = True if cropping else False
pic.csvheader(csvfile[n], cropped=cpd, pisascore=True)
if invals['OUTCSVPATH'][n] is not None:
if os.path.isfile(invals['OUTCSVPATH'][n]) is False:
pic.csvheader(invals['OUTCSVPATH'][n],
cropped=cpd,
pisascore=True)
logger.info('Parsing sequence files...')
for i, fpath in fseq.items():
# seq.imer[i].seqs['conkit'] = ckio.read(fpath, 'fasta')[0]
seq.imer[i].seqs['conkit'] = pio.read(fpath, 'fasta', ck=True)[0]
logger.info('Parsing contact predictions lists...')
conpred = {}
matches = []
for s in seq.imer:
if s not in conpred:
conpred[s] = {}
fs = fcropseq[s] if cropping else fseq[s]
for source, attribs in sources.items():
fc = os.path.splitext(os.path.basename(fs))[0]
fc += os.extsep + attribs[1]
confile = os.path.join(dmpdir, fc)
# conpred[s][source] = ckio.read(confile, attribs[2])[0]
conpred[s][source] = pio.read(confile, attribs[2], ck=True)[0]
logger.info('Parsing crystal structure contacts...')
for i in range(len(iflist)):
logger.info(os.linesep + str(iflist[i]))
fs = fcropstr if cropping else fstr
fs = (os.path.splitext(os.path.basename(fs))[0] + os.extsep +
"interface" + os.extsep + str(i + 1) + os.extsep + "pdb")
spath = os.path.join(pisadir, fs)
# inputmap = ckio.read(spath, 'pdb')
inputmap = pio.read(spath, 'pdb', ck=True)
if len(inputmap) == 4:
chnames = [
iflist[i].chains[0].crystal_id, iflist[i].chains[1].crystal_id
]
iflist[i].chains[0].seq_id = seq.whatseq(chnames[0])
iflist[i].chains[1].seq_id = seq.whatseq(chnames[1])
chseqs = [iflist[i].chains[0].seq_id, iflist[i].chains[1].seq_id]
logger.info(iflist[i].chains)
chtypes = [iflist[i].chains[0].type, iflist[i].chains[1].type]
if (chseqs[0] != chseqs[1]
or (chtypes[0] != 'Protein' or chtypes[1] != 'Protein')):
if chtypes[0] != "Protein" or chtypes[1] != "Protein":
logger.info(
'Interface ' + str(i) +
' is not a Protein-Protein interface. Ignoring.')
else:
logger.info('Interface ' + str(i) +
' is not a homodimer. Ignoring.')
iflist[i].structure = None
matches.append(None)
continue
s = chseqs[0]
try:
iflist[i].structure = []
for m in range(len(inputmap)):
iflist[i].structure.append(inputmap[m].as_contactmap())
iflist[i].structure[m].id = inputmap[m].id
except Exception:
logger.warning('Contact Maps obtained from a legacy ConKit ' +
'version with no Distograms implemented.')
for m in range(len(inputmap)):
iflist[i].structure.append(inputmap[m]) # ConKit LEGACY.
#fs = fcropstr if cropping else fstr
#fs = (os.path.splitext(os.path.basename(fs))[0] +
# os.extsep + "interface" + os.extsep + str(i+1) + os.extsep + "con")
#spath = os.path.join(pisadir, fs)
#pio.write(spath, 'psicov', indata=iflist[i].structure[1])
#iflist[i].contactmap = pio.read(spath, 'array')
iflist[i].contactmap = iflist[i].structure[1].deepcopy()
matches.append({})
for source, attribs in sources.items():
matches[i][source] = pcc.contact_atlas(
name=pdbid + '_' + str(s),
dimer_interface=iflist[i],
conpredmap=conpred[s][source],
conpredtype=source,
sequence=seq.imer[s])
if cropping is True:
matches[i][source].set_cropmap()
matches[i][source].remove_neighbours(mindist=2)
matches[i][source].set_conpred_seq()
matches[i][source].remove_intra()
matches[i][source].make_match(filterout=attribs[3])
for cmode, cmap in matches[i][source].conkitmatch.items():
if (len(cmap) > 0 and len(
matches[i][source].interface.structure[1]) > 0):
for imtype in plotformats:
if len(matches[i][source].conkitmatch) > 1:
pout = (os.path.splitext(fs)[0] + os.extsep +
'match' + os.extsep + cmode +
os.extsep + source + os.extsep +
'con' + os.extsep + imtype)
else:
pout = (os.path.splitext(fs)[0] + os.extsep +
'match' + os.extsep + source +
os.extsep + 'con' + os.extsep + imtype)
plotpath = os.path.join(
os.path.dirname(csvfile[0]), pout)
matches[i][source].plot_map_alt(plotpath,
mode=cmode,
plot_type=imtype)
else:
iflist[i].structure = None
iflist[i].contactmap = None
matches.append(None)
continue
logger.info(os.linesep + 'Computing results and writing them to file...' +
os.linesep)
for i in range(len(iflist)):
logger.info('Generating Interface ' + str(i + 1) + ' data...')
if matches[i] is None:
continue
results = [pdbid, str(i + 1)]
results.append(iflist[i].chains[0].crystal_id)
results.append(iflist[i].chains[1].crystal_id)
sid = iflist[i].chains[0].seq_id
results.append(str(sid))
results.append(str(seq.imer[sid].length()))
if cropping is True:
results.append(str(seq.imer[sid].cropmsa.meff))
else:
results.append(str(seq.imer[sid].msa.meff))
results.append(str(seq.imer[sid].ncrops()))
results.append(str(seq.imer[sid].full_length()))
results.append(str(seq.imer[sid].msa.meff))
for source, attribs in sources.items():
appresults = pcs.list_scores(matches[i][source], tag=source)
results.extend(appresults)
results.append(str(iflist[i].stable))
for n in range(2):
if scoring[n] is True:
pic.lineout(results, csvfile[n])
pic.lineout(results, invals['OUTCSVPATH'][n])
endmsg = pcl.ok(starttime, command=__script__)
logger.info(endmsg)
return
def main():
parser = create_argument_parser()
args = parser.parse_args()
global logger
logger = pcl.pisacov_logger(level="info")
welcomemsg, starttime = pcl.welcome(command=__script__)
logger.info(welcomemsg)
csvfile = ppaths.check_path(args.scores[0], 'file')
outdir = ppaths.check_path(args.outdir)
ppaths.mdir(outdir)
# Parsing scores
scores = {}
names = None
thraw = {}
with open(csvfile, 'r') as fin:
scoresin = csv.reader(fin)
for entry in scoresin:
if entry[0][0] != "#":
if (names is None or isinstance(names, str) or
(isinstance(names, list) and len(names) != len(entry))):
names = []
for n in range(len(entry)):
names.append('sc_' + str(n + 1))
else:
if thraw == {}:
for name in names[13:-1]:
thraw[name] = []
if entry[0] not in scores:
scores[entry[0]] = {}
if entry[1] not in scores[entry[0]]:
scores[entry[0]][entry[1]] = []
for sc in (entry.split(sep=', ')[13:-1]):
scores[entry[0]][entry[1]].append(float(sc))
if (entry.split(sep=', ')[-1]) == 'True' or '1':
scores[entry[0]][entry[1]].append(True)
elif (entry.split(sep=', ')[-1]) == 'False' or '0':
scores[entry[0]][entry[1]].append(False)
for n in range(len(names)):
thraw[names[n]].append(scores[entry[0]][entry[1]][n])
else:
if entry.split(
sep=', ')[13:] == scores[entry[0]][entry[1]]:
pass
else:
raise ValueError(
'CSV file contains different values for same interface.'
)
else:
names = entry[1:].split(sep=', ')
# Setting thresholds
thr = {}
FPR = {}
TPR = {}
for key, value in thraw.items():
thr[key] = list(set(thraw)).sort()
FPR[key] = []
TPR[key] = []
for t in thr[key]:
FP = 0
TP = 0
FN = 0
TN = 0
for pdbid in scores:
for iface in scores[pdbid]:
stable = scores[pdbid][iface][-1]
for n in range(len(names)):
if scores[pdbid][iface][n] < t:
if stable is True:
FN += 1
else:
TN += 1
else:
if stable is True:
TP += 1
else:
FP += 1
FPR[key].append(FP / (FP + TN))
TPR[key].append(TP / (TP + FN))
fnameout = os.path.join(
outdir,
(key + os.path.splitext(os.path.basename(csvfile))[0] + 'roc.dat'))
with open(fnameout, 'w') as fout:
for n in range(len(FPR[key])):
fout.write(str(FPR[key][n]) + ' ' + str(TPR[key][n]))
endmsg = pcl.ok(starttime, command=__script__)
logger.info(endmsg)
return