#!/usr/bin/env python # -*- coding: utf-8 -*- """Wrapper for ROSETTA/FARFAR2 (2020 update) software for structure prediction of small RNA sequences. Scores:: ------------------------------------------------------------ Scores Weight Raw Score Wghtd.Score ------------------------------------------------------------ fa_atr 0.210 -1081.212 -227.054 fa_rep 0.200 180.360 36.072 fa_intra_rep 0.003 1504.776 4.364 lk_nonpolar 0.250 -26.712 -6.678 fa_elec_rna_phos_phos 1.700 -3.955 -6.724 rna_torsion 1.000 63.080 63.080 suiteness_bonus 1.000 0.000 0.000 rna_sugar_close 0.820 13.043 10.696 fa_stack 0.130 -1144.488 -148.783 stack_elec 0.760 -17.443 -13.257 geom_sol_fast 0.170 355.629 60.457 hbond_sr_bb_sc 0.960 -12.355 -11.861 hbond_lr_bb_sc 0.960 -0.562 -0.539 hbond_sc 0.960 -76.482 -73.423 ref 1.000 257.400 257.400 free_suite 2.000 0.000 0.000 free_2HOprime 1.000 0.000 0.000 intermol 1.000 0.000 0.000 other_pose 1.000 0.000 0.000 loop_close 1.000 0.000 0.000 linear_chainbreak 5.000 0.000 0.000 --------------------------------------------------- Total weighted score: -56.252 protocols.rna.denovo.movers.RNA_Minimizer: RNA minimizer finished in 230 seconds. csv:: rna_csv.py --flat FARFAR2_hires-3.7.19+5.gd7c6fd7.dirty-mx.local.csv id 1 fn 1y26X_output4_01-000085_AA.pdb ff2_score_hires -56.252 ff2_fa_atr -227.054 ff2_fa_rep 36.072 ff2_fa_intra_rep 4.364 ff2_lk_nonpolar -6.678 ff2_fa_elec_rna_phos_phos -6.724 ff2_rna_torsion 63.08 ff2_suiteness_bonus 0.0 ff2_rna_sugar_close 10.696 ff2_fa_stack -148.783 ff2_stack_elec -13.257 ff2_geom_sol_fast 60.457 ff2_bond_sr_bb_sc -11.861 ff2_hbond_lr_bb_sc -0.539 ff2_hbond_sc -73.423 ff2_ref 257.4 ff2_free_suite 0.0 ff2_free_2HOprime 0.0 ff2_intermol 0.0 ff2_other_pose 0.0 ff2_loop_close 0.0 ff2_linear_chainbreak_hires 0.0 """ import os, shutil, re import subprocess import os from shutil import copyfile from rna_tools.tools.mq.lib.wrappers.SubprocessUtils import run_command from rna_tools.tools.pdb_formatix.PDBFile import PDBFile#resname_check_and_3to1, set_residues_bfactor from rna_tools.tools.mq.lib.wrappers.base_wrappers import ProgramWrapper, WrapperError from rna_tools.tools.pdb_formatix.RosettaUtils import RosettaPDBFile from rna_tools.tools.pdb_formatix.rebuilder import check_and_rebuild from rna_tools.rna_tools_config import FARFAR2_PATH, FARFAR2_DB_PATH, FARFAR2_LORES class FARFAR2(ProgramWrapper): """ Wrapper class for running ROSETTA scoring function automatically. """ program_name = 'farna' src_bin = FARFAR2_PATH db_path = FARFAR2_DB_PATH input_fn = 'seq.fasta' input_file = '' best_energy = '' executable = FARFAR2_PATH def __init__(self, job_id=None): try: self.start_dir = os.getcwd() except OSError: # directory was deleted or something like that pass super(FARFAR2, self).__init__('', '', job_id=job_id) def _prepare_stderr_stdout(self): # create output file self.output_file = os.path.join(self.path, 'stdout.txt') self.stdout = open(self.output_file, 'w') # create error file self.error_file = os.path.join(self.path, 'stderr.txt') self.stderr = open(self.error_file, 'w') def _prepare_files(self): # create input file self.input_file = open(os.path.join \ (self.sandbox_dir, self.input_fn), 'w')\ .write('>seq.fasta\n'+str(self.sequence).lower()) self._prepare_stderr_stdout() def sandbox(self): #shutil.copytree(self.src_bin + os.sep + 'rosetta_source', # self.sandbox_dir + os.sep + 'rosetta_source', # symlinks=True) #os.symlink(self.src_bin, self.sandbox_dir + os.sep + self.executable) #symlinks=True) #os.symlink(FARNA_DB_PATH, # self.sandbox_dir + os.sep + 'rosetta_database') #os.system('chmod +x %s' % \ # os.path.join(self.sandbox_dir, self.executable)) pass def run(self, pdb_file, hires, verbose=False):#, global_energy_score=True): """Compute FARNA potential for a single file Arguments: * pdb_file = path to pdb file * global_energy_score = True/False (See Output), default=True Output: * A list of energies, e.g:: ['-21.721', '-0.899', '-20.961', '-84.498', '-16.574', '-180.939', '11.549', '7.475', '-17.257', '-306.324', '0.0', '0.0', '17.503', '0.0'] ??? or a dictionary of lists of local scores, eg:: { 'N_BS': [17.0, -0.70039, -0.720981, -0.685238, -0.734146, ... ], 'atom_pair_constraint': [0.0, -0.764688, -0.773833, ...], ... } """ global_energy_score=True ftxt = open(pdb_file).read() ftxt = re.sub('TER\s+END\s+', 'TER', ftxt) ftxt = re.sub('END', 'TER', ftxt).strip() f = open(self.sandbox_dir + os.sep + 'tmp.pdb', 'w') f.write(ftxt) f.close() pdb_file = self.sandbox_dir + os.sep + 'tmp.pdb' if check_and_rebuild(pdb_file, self.sandbox_dir + os.sep + 'query.pdb'): self.pdb_fixes.append('rebuild_full_atom') pdb_file = RosettaPDBFile(pdb_path=self.sandbox_dir + os.sep + 'query.pdb') # get sequence from PDB file with open(self.sandbox_dir + os.sep + 'query.fasta', 'w') as f: f.write(pdb_file.get_fasta(lowercase=True)) # create a ROSETTA ready PDB file pdb_file.make_rna_rosetta_ready() pdb_file.save(self.sandbox_dir + os.sep + 'query.pdb') self.pdb_fixes = pdb_file.fixes # run #os.chdir(self.sandbox_dir) self.flags = [self.sandbox_dir + os.sep + self.executable] if hires == False: # must be a string minimize_cmd = ' -score:weights ' + FARFAR2_LORES + ' ' # # -minimize_rna' ## # MM minimize_rna should be off or by option else: minimize_cmd = ' ' # self.sandbox_dir + os.sep + # verbose = True cmd = ' '.join([self.executable , '-database', self.db_path, minimize_cmd, ' -ignore_zero_occupancy false -constant_seed ', '-s', self.sandbox_dir + os.sep + 'query.pdb', '-out:file:silent', self.sandbox_dir + os.sep + 'SCORE.out']) self.log(cmd, 'debug') if verbose: print(cmd) self.log('Running program') system = False # for debuggings if system: os.system(cmd) else: out = subprocess.getoutput(cmd) self.log('Run finished') self.log(out, 'debug') self.get_result() #if global_energy_score: # ??? results = [] for i in list(self.result.keys()): results.append(float(self.result[i][0])) # float vs string #return '\t'.join(results) return results def get_result(self): """Parse and get result from score file created during ROSETTA run All results are kept in self.result, but only global score is returned """ try: with open(self.sandbox_dir + os.sep + 'SCORE.out') as f: lines = f.readlines() #().split('\n') except: raise Exception('Problem with gettting the results, run with --verbose') lines = [l for l in lines if not l.startswith('REMARK')] #print(lines) # get names of different scores keys = lines[1].split()[1:-1] # get global scores global_scores = lines[2].split()[1:] # global scores are at index 0 in result, local are at 1--len(sequence) self.result = dict(list(zip(keys, [[float(s)] for s in global_scores[:len(keys)]]))) ##for l in lines[3:-1]: # scores_res = l.split()[2:-1] # scores for a single residue # for i in xrange(len(keys)): # self.result[keys[i]].append(float(scores_res[i])) #return self.result['score'][0] return self.result def mqap(self, pdb): "Total weighted score:\s+(?P[-\d.]+)" pass def cleanup(self): super(FARFAR2, self).cleanup() # main if __name__ == '__main__': fns = ['test.pdb'] fns = ['1xjrA_M1.pdb', 'test.pdb'] fns = ['3e5f_output4_01-000001_AA+ResnShift.pdb'] #2pcw_1_2chains.pdb'] # two chains for f in fns: f = 'test' + os.sep + f print('processing %s' % f) if 1: # mini false farna = FARFAR2() result = farna.run(f, False, verbose=True) print(result) if 1: # mini true farna = FARFAR2() result = farna.run(f, True, verbose=True) print(result) #farna.cleanup()