def funcpred_analysis(self): cursor = self.conn.cursor(MySQLdb.cursors.DictCursor) #kdrew: get domains from diff tables (eg. hddb_IEA_mygoLite_062005_062009) domains = NewAnnotatedDomains(self.conn, self.sequence_table, self.domain_table, self.evidence_codes) domains.load_domains() for domain in domains: #kdrew: if single domain flag is set and parent and domain sequence keys do not match, move to the next domain #kdrew: generally when parent sequence key and domain sequence key are not the same the protein is a single domain protein if self.single_domain and domain.get_parent_sequence_key() != domain.get_domain_sequence_key(): continue print domain.get_parent_sequence_key(), domain.get_domain_sequence_key() #kdrew: get_old_annotations old_anns = Terms(ecode=self.evidence_codes) old_anns.load_terms(conn=self.old_mg.get_conn(), psk=domain.get_parent_sequence_key(), go_type=self.term_types, term_table=self.old_annotations_table) #print "old annotations: " #old_anns.print_terms() #kdrew: get_new_annotations new_anns = Terms(ecode=self.evidence_codes) new_anns.load_terms(self.new_mg.get_conn(), domain.get_parent_sequence_key(), self.term_types, term_table=self.new_annotations_table) #print "new annotations: " #new_anns.print_terms() #kdrew: get_predictions preds = Predictions(conn=self.pred_conn, table=self.prediction_table) preds.load_predictions(domain.get_parent_sequence_key(), domain.get_domain_sequence_key()) #kdrew: predictions_trim = predictions - old_annotations predictions_trim = Predictions() #kdrew: produces a list of every prediction that is not in the old annotations predictions_trim.extend(list((x) for x in preds if x.function not in old_anns)) #print "predictions_trim: ", predictions_trim.filter(pls_llr=0, base_llr=-2) #kdrew: new_annotations_trim = new_annotations - old_annotations #kdrew: produces a list of only the new annotations that are truely new, not been seen before new_annotations_trim = list(set(new_anns) - set(old_anns)) #print "new annotations trim: ", new_annotations_trim #kdrew: predictions_trim & new_annotations_trim new_ann_trim_tmp = list((x) for x in new_annotations_trim if predictions_trim.getByAcc(x.acc)) #print "new_ann_trim_tmp: ", new_ann_trim_tmp prediction_new_annotation_intersection = Predictions() prediction_new_annotation_intersection.extend(list((x) for x in predictions_trim if x.function in new_annotations_trim)) if len(prediction_new_annotation_intersection) > 0: print "pred & new: " , prediction_new_annotation_intersection.filter(pls_llr=self.pls_min, base_llr=self.base_max)
