Diagnosis of disease-related genes is one of the most important goals in biomedical researchers. The topic is the arespredic-tion of disease genes which the genes associated with them still remain unknown. Most network-based method together.gbased on the principle that genes associated with diseases that are similar in phenotypes, are usually interactin enegThe main goal is to find these interactions or modules more accurate. Some of these approaches are based on gene esnetworks to discover new genes associated with the disease. Other methods integrate gene-gene networks with disea the,similarity network to improve the accuracy of discovering new genes. Despite the different integration approaches suproblem of accurately recognizing disease causing genes is still a challenge. Given the definition of heterogeneo egrationtnetworks made up of nodes with different kinds of connections, it is possible to consider a network made up of in inding afof different biological networks as a heterogeneous network. On the other hand, there are various algorithms for gmodule or cluster on heterogeneous networks such as RWRHN. Therefore, due to the existence of effective clusterin llyaalgorithms on heterogeneous networks, the problem of discovering the genes associated with the disease, which usu . form a cluster, can be examined from the perspective of cluster detection on the heterogeneous network. The method introduced in this research is called RWRHN-FF. First, four gene-similarity networks based on different ase,hgenomic resources were constructed. Then the type II fuzzy voter was used to combine the networks. In the second p ademthe gene-gene network used a two-part disease-gene network, was connected to a disease-disease similarity network Nup of a combination of four different sources, and a reliable heterogeneous network was created. Finally, the RWRH ultssalgorithm was implemented on a network constructed to prioritize the candidate gene for inherited diseases. The re show that the performance of RWRHN-FF is better than RWRHN-RE and RWRH. In addition, RWRHN-FF was used to mpredict new genes for prostate, breast, stomach and intestine cancers. Also, in order to reduce the RWRHN algorith runtime on heterogeneous networks, Apache Spark has been used, which converges in less time compared to other . implementations on different volumes of heterogeneous networks. Key Words :Clustering, Heterogeneous networks, Gene-Gene network, Genes associated with the disease, Prioritization, Type II Fuzzy Voter, RWRHN