interchiba / pdmda

======================================================================================================= Welcome to use PDMDA algorithm. PDMDA is a novel and effective computational model for predicting deep-level miRNA-disease association.

Requirements

PyTorch 1.1.0

Author:Cheng Yan and Jianxin Wang

E-mail:[email protected] [email protected]

======================================================================================================= PDMDA_task1.py: the 5-fold cross validation of predicting association type between known 4 association type samples PDMDA_task2.py: the 5-fold cross validation of predicting type between known 4 association type samples and un-type samples PDMDA_task3.py: the 5-fold cross validation of predicting the type between known 4 association type samples, untype samples and un-ass samples

###data/miRNAdisease-HMDD2.0-task1 directory CVdisease_types.txt:the diseases id of dataset (known 4 association type samples) CVmiRNA_types.txt:the miRNAs id of dataset (known 4 association type samples) label_types.npy: the association type labels of miRNA-disease pairs of dataset (known 4 association type samples) HMDD2_new_miRNAseqfeature.txt: the features of miRNAs diseasefeature-r.zip: the raduis feature of disease (unzip it to diseasefeature-r.npy) disease_list.txt: the uniquely diseases id of dataset fingerprint_dict.pickle: the uniquely gene dict of dataset disease_gene.pickle: the disease-gene interactions geneintmat.zip: the gene-gene interaction (unzip it to geneintmat.txt)

###data/miRNAdisease-HMDD2.0-task2 directory CVdisease_all_and_un_types.txt:the diseases id of dataset (known 4 association type samples and un-type samples) CVmiRNA_all_and_un_types.txt:the miRNAs id of dataset (known 4 association type samples and un-type samples) label_all_and_un_types.npy: the association type labels of miRNA-disease pairs of dataset (known 4 association type samples and un-type samples) HMDD2_new_miRNAseqfeature.txt: the features of miRNAs diseasefeature-r.zip: the raduis feature of disease (unzip it to diseasefeature-r.npy) disease_list.txt: the uniquely diseases id of dataset fingerprint_dict.pickle: the uniquely gene dict of dataset disease_gene.pickle: the disease-gene interactions geneintmat.zip: the gene-gene interaction (unzip it to geneintmat.txt)

###data/miRNAdisease-HMDD2.0-task3 directory CVdisease_all_and_un_ass.txt:the diseases id of dataset (known 4 association type samples, untype samples and un-ass samples) CVmiRNA_all_and_un_ass.txt:the miRNAs id of dataset (known 4 association type samples, untype samples and un-ass samples) label_all_and_un_ass.npy: the association type labels of miRNA-disease pairs of dataset (known 4 association type samples, untype samples and un-ass samples) HMDD2_new_miRNAseqfeature.txt: the features of miRNAs diseasefeature-r.zip: the raduis feature of disease (unzip it to diseasefeature-r.npy) disease_list.txt: the uniquely diseases id of dataset fingerprint_dict.pickle: the uniquely gene dict of dataset disease_gene.pickle: the disease-gene interactions geneintmat.zip: the gene-gene interaction (unzip it to geneintmat.txt)