The java application, jp.ac.kyushuu.math.om.SmallSizedComplex.NBCForHeterodimericComplexes, can be found in NBC-HD-Complexes.jar. Options: --out_dir specifies an output directory. --interactions specifies an input PPI file. --complexes specifies a file of known protein complexes. --ontology specifies the string representing a directory holding sim38_BP.txt and sim38_MF.txt files, which can be found in data_files.zip, followed by "/sim38" (slash and a prefix of those files). For example, if your home directory has sim38_BP.txt and sim38_MF.txt files, specify --ontology=~/sim38. --random_walk_threshold --random_walk_restart_prob --random_walk_tmp_dir specifies a directory of temporary files generated by random walks with restarts. --out_prefix specifies a prefix of the name of an output file. --features specifies features --fold --bins --pseudo_count --lower_bound specifies lower bound on the weights of PPIs. --localization specifies a source data file of the localization feature. --llr_threshold --neg_ex_ratio specifies the ratio of the number of negative examples to that of positive examples. --sgd_features SGD_features.tab The latest version of SGD features is http://downloads.yeastgenome.org/curation/chromosomal_feature/SGD_features.tab Usage example: java -Xms1024m -Xmx1024m -classpath NBC-HD-PCP.jar jp.ac.kyushuu.math.om.SmallSizedComplex.NBCForHeterodimericComplexes --out_dir=output --interactions=Kiemer2007wi-phi_3_s.txt --complexes=CYC2008_complex.txt --ontology=sim38 --random_walk_threshold=0.0000000001 --random_walk_restart_prob=0.6 --random_walk_tmp_dir=C:/local/var --out_prefix=output --features=features.txt --fold=5 --bins=10 --pseudo_count=1 --lower_bound=6 --localization=//thyme/om/data/yeastgfp.yeastgenome.org/allOrfData.txt --llr_threshold=0.6 --neg_ex_ratio=10 --sgd_features=SGD_features.tab.txt Once, the application is executed, the sets of positive and negative examples used in the learning process are saved as files in the specified output directory, named positiveExamples.txt and negativeExamples.txt, respectively. By the following options, they are loaded and used as positive and negative examples: --pos_examples= --neg_examples=