Weighted Benjamini-Hochberg over overlapping neighbourhoods

Milo's spatial FDR. Neighbourhoods overlap, so the tests are not independent and a plain BH is anti-conservative. Each p-value is weighted by the reciprocal of its connectivity and the step-up runs on those weights. Non-finite p-values are carried through untouched and take no part in the adjustment.

Usage

rs_spatial_fdr(p_values, connectivity)

Arguments

p_values

Numeric vector. One raw p-value per tested neighbourhood.

connectivity

Numeric vector. The matching connectivity per neighbourhood, either the k-th neighbour distances or the neighbourhood overlaps. A zero connectivity gets a weight of one, as in the upstream.

Value

The adjusted p-values, in the input order.

References

Dann, et al., Nat Biotechnol, 2022