Permutation p-value approximation via generalized Stolarsky invariance

TL;DR

This paper introduces a fast, geometry-based approximation method for permutation p-values in genomic data analysis, improving efficiency and accuracy over existing saddlepoint methods.

Contribution

It develops a novel approximation for permutation p-values using Stolarsky's invariance principle, with variance estimation, applicable to two-sample linear test statistics.

Findings

The method provides accurate p-value estimates with modest variance.

It outperforms saddlepoint approximations in speed and accuracy on Parkinson's data.

The approach offers a probabilistic interpretation of Stolarsky's invariance principle.

Abstract

It is common for genomic data analysis to use $p$-values from a large number of permutation tests. The multiplicity of tests may require very tiny $p$-values in order to reject any null hypotheses and the common practice of using randomly sampled permutations then becomes very expensive. We propose an inexpensive approximation to $p$-values for two sample linear test statistics, derived from Stolarsky's invariance principle. The method creates a geometrically derived set of approximate $p$-values for each hypothesis. The average of that set is used as a point estimate $\hat p$ and our generalization of the invariance principle allows us to compute the variance of the $p$-values in that set. We find that in cases where the point estimate is small the variance is a modest multiple of the square of the point estimate, yielding a relative error property similar to that of saddlepoint approximations. On a Parkinson's disease data set, the new approximation is faster and more accurate than the saddlepoint approximation. We also obtain a simple probabilistic explanation of Stolarsky's invariance principle.