The Cauchy Combination Test under Arbitrary Dependence Structures

Combining individual p-values to perform an overall test is often encountered in statistical applications. The Cauchy combination (CCT) (Journal of the American Statistical Association, 2020, 115, 393–402) a powerful and computationally efficient approach integrate under arbitrary dependence structures for sparse signals. We revisit this additionally show that (i) tail probability CCT can be approximated just as well when more relaxed assumptions are imposed on compared those original statistics; (ii) such satisfied by six popular copula distributions; (iii) power no less than minimum p-value number goes infinity some regularity conditions. These findings confirmed both simulations applications two real datasets, thus, further broadening theory CCT.