CRESTA: a comprehensive transcriptome atlas for cellular response to external stressors

Abstract

Cellular stress response (CSR) is crucial for maintaining intracellular homeostasis upon exposure to hazardous environmental stressors, whose failure can lead to cell death. Here, we developed CRESTA (https://cresta.renlab.cn/), an integrated database providing a comprehensive resource for studying stressor-induced CSR. CRESTA unifies fragmented transcriptomic data across diverse stressors, cell types, and contexts into a hierarchically structured atlas. It catalogs 20 786 unique stress-associated genes responding to 14 major categories (180 sub-categories) of human cellular stressors, based on the differential expression analysis of 8258 samples spanning 197 human cell types. These stressor categories include air pollutants, antineoplastic agents, heavy metals, hypoxia, mechanical stimuli, natural toxins, nutrient deprivation, pesticides, radiation, temperature change, etc. To further enable causal inference between stressors and molecular pathologies, CRESTA links CSR transcriptomic signatures to functional pathways, cell death annotations, and disease associations. Additionally, the CSR profiles are connected with chemical perturbation features involving FDA-approved drugs to support systematic drug repurposing. Overall, we expect that CRESTA will serve as a vital resource for mechanistic studies of cytotoxicity, cellular perturbation evaluation, and CSR-targeted drug discovery.