Orthology-Aware Reanalysis of Replicated Public Transcriptomes Identifies Shared and Context-Dependent Low-Temperature Responses in Teleost Fishes.

Low temperature constrains the survival and production of many cultured fishes, yet public cold-response transcriptomes differ substantially in species, developmental stage, tissue, exposure design, and genome annotation. We therefore reassessed the extent to which replicated public datasets support shared rather than universal low-temperature responses. A metadata screen identified 110 candidate BioProjects containing 1,465 runs. After manual verification of treatment relevance, sequencing platform, biological design, and read depth, 15 projects (208 runs; 10 species) were retained as a public-data landscape. Only datasets with at least three biological replicates per group and traceable processed abundance tables entered differential-expression inference: Nile tilapia (Oreochromis niloticus) adult midbrain (four cold and four control samples) and zebrafish (Danio rerio) whole larvae (three cold and three matched controls at five time points). Differentially expressed genes were defined by false discovery rate (FDR) < 0.05 and |log2 fold change| >= 1. Unbiased Gene Ontology and Kyoto Encyclopedia of Genes and Genomes analyses identified recurrent rhythmic/transcriptional, MAPK-FoxO stress-signaling, cell-fate, innate-immune, and metabolic themes. Ensembl one-to-one orthology restricted the primary cross-species comparison to 6,383 expression-mapped gene pairs. Thirty-two pairs showed concordant differential expression in tilapia and zebrafish at 24 h, whereas six were discordant. Candidate counts ranged from 10 to 91 across prespecified FDR and effect-size thresholds. Replicated common carp (Cyprinus carpio) brain and heart results were used as an external, paralogue-aware check: 11 of the 32 pairs received same-direction support, three showed an opposite response, and 18 lacked strong support. nr1d1, bhlhe40, dusp1, plin2, pprc1, and pfkfb3 illustrate conserved primary responses together with tissue- and lineage-dependent divergence. These associations generate focused hypotheses for experimental testing but do not establish causal cold-tolerance genes.