gLinDA: global differential abundance analysis ofmicrobiomes

Abstract

Microbiome composition plays a significant role in various diseases, including cancer, obesity, inflammatory bowel disease, and mental health disorders. Understanding the differences in microbial abundance between patients is key to uncovering the microbiomes impact on these conditions. Differential abundance analysis (DAA) can detect significant changes in microbiomes at the taxa level base. However, since individuals have unique microbial fingerprints that can potentially identify them, microbiome data must be treated as sensitive. We introduce gLinDA, a global differential abundance analysis software offering a swarm learning approach to analyzing distributed datasets while maintaining result accuracy without sharing patients raw data.