The study performed a comprehensive computational analysis of the Arabidopsis thaliana proteome, classifying 48,359 proteins by melting temperature (Tm) and melting temperature index (TI) and linking thermal stability to amino acid composition, molecular mass, and codon usage. Machine‑learning and evolutionary analyses revealed that higher molecular mass and specific codon pairs correlate with higher Tm, and that gene duplication has driven the evolution of high‑Tm proteins, suggesting a genomic basis for stress resilience.

The study demonstrates that abscisic acid (ABA) accumulates in darkness to suppress cotyledon opening during seedling deetiolation, and that light exposure lifts this repression, enabling cotyledon aperture. Genome‑wide transcriptional and alternative‑splicing changes accompany this process, and the light‑dependent regulation requires the splicing factors RS40 and RS41, whose activity is repressed in the dark.