Back

TestNet: A Testing Method for Inferring Microbial Networks with False Discovery Rate Control, for Clustered and Unclustered Samples

Su, C.; He, M.; Van Doren, V. E.; Kelley, C. F.; Hu, Y.

2025-04-19 bioinformatics
10.1101/2025.04.14.648862 bioRxiv
Show abstract

Most existing methods for inferring microbial networks generate only point estimates of Pearsons correlations without assessing their significance, and none can account for clustered samples. In this article, we introduce TestNet, a novel method that delivers well calibrated results by controlling the false discovery rate (FDR). We developed a permutation-based procedure to generate valid null replicates that account for compositional effects and extensive zeros in microbiome data, and clustering structures within the samples when present. Our results demonstrate that TestNet is the only method that effectively controls the FDR while maintaining high power across a wide range of scenarios.

Published in Genome Biology (predicted rank #5) · training set

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.