Identification and Validation of a Prognostic Risk-Scoring Model Based on LATS2 Expression in Acute Myeloid Leukemia.

Acute myeloid leukemia (AML)'s treatment and remission remains unsatisfactory. A prognostic risk-scoring model containing seven signature genes (POU3F1, RPGR, PTP4A3, SOCS1, FAM83G, GREB1 and COL2A1), was developed by LASSO-Cox regression analysis. In the training set, the test group, area under the curve values of 1, 3, and 5 years were 0.876, 0.877, 0.937, and 0.974, 0.878, 0.976 respectively, which indicates a good predictive efficacy. In the two external GEO (GSE71014 and GSE6891) datasets, area under the curve values of 1, 3, 5 years were 0.847, 0.857, 0.822, and 0.830, 0.863, 0.891 respectively. Our seven signature genes containing risk-scoring model performed excellently in evaluating the OS of AML patients.