Estimating Proportion of True Null Hypotheses based on Sum of p-values and application to microarrays
share
A new estimator of proportion of true null hypotheses based on sum of all p- values has been proposed in this work which removes the problem of choosing tuning parameters in the existent estimators. Normality of gene expression levels and common t-test for each gene are assumed as in some significant recent works. Simulation studies show that the proposed estimator performs better than its closest competitor over an important continuous sub-interval of the parameter space under weak dependence among the gene expression levels.
READ FULL TEXT