Skip to main content
ISMPMI Congress 2025 Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.

Concurrent Session

Session: Concurrent Session: Plant Regulation of biotic and abiotic stress – trade-offs

HIVE: a novel unsupervised machine learning tool to find specific and common plant defence signatures against multi-factorial stresses

Thursday, July 17, 2025
11:00 - 11:10  
Location: Room P+Q

Abstract Text:

Plants are subjected to multiple concomitant stresses both biotic and abiotic. Studying the response of one organism to simultaneous multi-factorial stresses is complex. Current methods to integrate unpaired omics data that study these stresses separately, try to overcome this complexity but often lead to a prioritization of specific response signatures over the common ones. For this purpose, we developed HIVE (Horizontal Integration analysis using Variational AutoEncoders), a novel computational tool to integrate and analyse single stress unpaired transcriptomics data to identify specific and multi-stress responding signatures. HIVE couples a variational autoencoder, to alleviate batch effects, and uses a random forest regression with SHAP explainer to select relevant genes modulated in response to one or multiple stresses. We illustrate the functionality of HIVE using microarray and RNA-sequencing publicly available data coming from five important crop species and the model organism Arabidopsis thaliana. We showed that HIVE, applied to transcriptomics data, overcomes state-of-the-art methods and identifies novel promising candidates of effective defence responses to multi-factorial stresses, leading to the unprecedented validation in planta of two NBS-LRR genes in one of the analysed crop species.