With　the　rapid　growth　of　microarray　data,　it　has　become　a　hot　topic　to　reveal　complex　behaviors　and　functions　of　life　system　by　studying　the　relationships　among　genes.　In　the　process　of　reverse　network　modeling,　the　relationships　with　less　relevance　are　generally　not　considered　by　determining　a　threshold　when　the　relationships　among　genes　are　mined.　However,　there　are　no　effective　methods　to　determine　the　threshold　up　to　now.　It　is　worthwhile　to　note　that　the　phenotypes　of　genetic　diseases　are　generally　regarded　as　external　representation　of　the　functions　of　genes.　Therefore,　two　types　of　phenotype　networks　are　constructed　from　gene　and　disease　views,　respectively,　and　through　comparing　these　two　types　of　phenotype　networks,　the　threshold　of　gene　network　corresponding　to　a　certain　disease　can　be　determined　when　their　similarity　reaches　to　maximum.　Because　the　gene　network　is　determined　based　on　the　relationships　among　phenotypes　and　phenotypes　are　external　representation　of　the　functions　of　genes,　it　is　considered　that　relationships　in　the　gene　network　may　show　functional　relationships　among　genes　in　biological　system.　In　this　work,　the　thresholds　0.47　and　0.48　of　gene　network　are　determined　based　on　Parkinson　disease　phenotypes.　Furthermore,　the　validity　of　these　thresholds　is　verified　by　the　specificity　and　susceptibility　of　phenotype　networks.　Also,　through　comparing　the　structural　parameters　of　gene　networks　for　normal　and　disease　stage　at　different　thresholds,　significant　difference　between　the　two　gene　networks　at　threshold　0.47　or　0.48　is　found.　The　significant　difference　of　structural　parameters　further　verifies　the　efficiency　of　this　method.