IncRNA-disease association prediction method based on high-order proximity and matrix completion algorithm

The invention provides an lncRNA-disease association prediction method based on high-order proximity and a matrix completion algorithm. The lncRNA-disease association prediction method comprises the following steps: S1, calculating a high-order proximity matrix of an lncRNA similar matrix LS and a disease similar matrix DS; s2, acquiring a disease-lncRNA adjacency matrix DL, wherein the disease-lncRNA adjacency matrix DL is used for describing an lncRNA-disease association relationship; s3, constructing a heterogeneous disease-lncRNA incidence matrix, wherein the disease-lncRNA incidence matrix integrates a disease-lncRNA adjacent matrix DL, a high-order proximity matrix of a lncRNA similar matrix LS and a high-order proximity matrix of a disease similar matrix DS; and S4, predicting the association of the lncRNA and the disease in the disease-lncRNA association matrix by adopting a matrix completion method. According to the method, high-order proximity is introduced to reconstruct a similarity matrix of the