| """ |
| preprocess for algorithm. |
| For example, you generally need to convert AF3 input JSON files to the algorithm expected format. |
| You can use PreProcess.preprocess() to perform preprocessing. |
| """ |
|
|
| import argparse |
| import os |
| import json |
| from tqdm import tqdm |
|
|
|
|
| class PreProcess(): |
| def __init__(self): |
| super().__init__() |
| |
| def format_single_input(self, input_data): |
| tmp_seq_data={} |
| tmp_seq_data['name']= input_data['name'] |
| tmp_seq_data["sequences"]=[] |
|
|
| for seq in input_data['sequences']: |
| tmp_seq={} |
|
|
| seq_type = next(iter(seq)) |
|
|
| if seq_type=='protein': |
| tmp_seq["proteinChain"]={} |
| tmp_seq["proteinChain"]['sequence']=seq[seq_type]['sequence'] |
| if isinstance(seq[seq_type]['id'],list): |
| tmp_seq["proteinChain"]['count']=len(seq[seq_type]['id']) |
| else: |
| tmp_seq["proteinChain"]['count']=1 |
| |
| |
| |
| tmp_seq["proteinChain"]['modifications']=[] |
| for modification in seq[seq_type]['modifications']: |
| tmp_modification = {} |
| tmp_modification["ptmType"]=f'CCD_{modification["ptmType"]}' |
| tmp_modification['ptmPosition']=modification['ptmPosition'] |
| tmp_seq["proteinChain"]['modifications'].append(tmp_modification) |
| |
|
|
| elif seq_type=='rna': |
| tmp_seq["rnaSequence"]={} |
| tmp_seq["rnaSequence"]['sequence']=seq[seq_type]['sequence'] |
| if isinstance(seq[seq_type]['id'],list): |
| tmp_seq["rnaSequence"]['count']=len(seq[seq_type]['id']) |
| else: |
| tmp_seq["rnaSequence"]['count']=1 |
| |
| |
| tmp_seq["rnaSequence"]['modifications']=[] |
| for modification in seq[seq_type]['modifications']: |
| tmp_modification = {} |
| tmp_modification["modificationType"]=f'CCD_{modification["modificationType"]}' |
| tmp_modification['basePosition']=modification['basePosition'] |
| tmp_seq["rnaSequence"]['modifications'].append(tmp_modification) |
|
|
| elif seq_type=='dna': |
| tmp_seq["dnaSequence"]={} |
| tmp_seq["dnaSequence"]['sequence']=seq[seq_type]['sequence'] |
| if isinstance(seq[seq_type]['id'],list): |
| tmp_seq["dnaSequence"]['count']=len(seq[seq_type]['id']) |
| else: |
| tmp_seq["dnaSequence"]['count']=1 |
| |
| |
| tmp_seq["dnaSequence"]['modifications']=[] |
| for modification in seq[seq_type]['modifications']: |
| tmp_modification = {} |
| tmp_modification["modificationType"]=f'CCD_{modification["modificationType"]}' |
| tmp_modification['basePosition']=modification['basePosition'] |
| tmp_seq["dnaSequence"]['modifications'].append(tmp_modification) |
| |
| elif seq_type=='ligand': |
| tmp_seq["ligand"]={} |
| |
| tmp_seq["ligand"]['ligand']="CCD" |
| for ccd in seq[seq_type]['ccdCodes']: |
| tmp_seq["ligand"]['ligand'] += f'_{ccd}' |
| |
| if isinstance(seq[seq_type]['id'],list): |
| tmp_seq["ligand"]['count']=len(seq[seq_type]['id']) |
| else: |
| tmp_seq["ligand"]['count']=1 |
|
|
| tmp_seq_data["sequences"].append(tmp_seq) |
|
|
| return tmp_seq_data |
|
|
| def preprocess(self, af3_input_json_path, input_dir): |
| """ |
| af3_input_json_path: path to the af3 input json file |
| input_dir: path to the input directory for the algorithm |
| """ |
|
|
| with open(af3_input_json_path, "r") as f: |
| folding_inputs = json.load(f) |
| |
| |
| mapped_folding_inputs = [ |
| self.format_single_input(folding_inputs[i]) |
| for i in tqdm(range(len(folding_inputs))) |
| ] |
| |
| with open(f'{input_dir}/inputs.json', "w") as f: |
| json.dump(mapped_folding_inputs, f,indent = 4) |
| print( |
| "{} folding inputs written to {}.".format(len(mapped_folding_inputs), input_dir) |
| ) |
| |
|
|
| parser = argparse.ArgumentParser() |
| parser.add_argument( |
| "--af3_input_json", |
| help="The path to the input .json file.", |
| ) |
| parser.add_argument( |
| "--input_dir", |
| help="The path to write prepared input data in the format expected by the algorithm.", |
| ) |
| args = parser.parse_args() |
|
|
| preprocess = PreProcess() |
| preprocess.preprocess(args.af3_input_json,args.input_dir) |
|
|
