Integrating omic data to detect association between sHSP transcripts and phenotypic variability in tomato fruits

CACCHIARELLI, PAOLO; SPETALE, F.; TAPIA, E.; GUILLERMO RAUL PRATTA

Rosario, Santa Fe

Congreso; XIII Congreso Argentino de Bioinformática y Biología Computacional, XIII Conferencia Internacional de la Sociedad Iberoamericana de Bioinformática (XIII SoIBio) y III Reunión Anual de la Red Iberoamericana de Inteligencia Artificial para Big BioData (III; 2023

SoIBio, RIABio, A2B2C

Fruit ripening is a complex developmental process highly coordinated by different gene families such as small Heat Shock Proteins (sHSP), witch maintain cell homeostasis and are a superfamily of chaperones that have been characterized in other organisms. In plants, sHSP promote protein folding and disaggregation during stress or developmental changes. Previously, we obtained the transcriptome of cv. Caimanta (C, S. lycopersicum), the exotic LA0722 (P, S. pimpinellifolium), and their interspecific hybrid (CxP) in 3 main stages of ripening. Also, agronomic characterization of fruit attributes was achieved in the three genotypes. The objective of this communication was to evaluate the association between both datasets. A Generalized Procrustean Analysis (GPA) was applied to gene expression levels of two clusters of sHSP located on chromosomes 6 and 9, respectively, as transcriptomic data, and 11 quantitative fruit traits as phenomic data. Also, an estimation of the degrees of dominance (d/a) was carried out for the levels of genes expression and the fruit traits. Principal Components 1 (PC1) and 2 (PC2) from the GPA explained the 77.4% and the 22,6% of the total variability, respectively. In the biplot, the consensus positions of P and CxP were close at positive values of PC1, while C located on the negative values. Contrarily, P and CxP were discriminated by PC2 because the exotic parent was on its negative values and the hybrid on the positive ones. C was located at 0 value. Nevertheless, for each genotype positions according to transcriptomic and phenomic characterizations were similar, indicating a high association among both datasets. Also, d/a for all traits evidenced a high dominance of the wild genome, which agreed to the proximity of P and CxP in the biplot. Concluding, the integration of transcriptomic and phenomic data by PGA and d/a estimation allow identifying a high degree of association amongexpression levels of both sHSP genes and quantitative traits. This information can be applied in breeding programs to get an optimal use of exotic genes in obtaining new varieties with adequate fruit quality traits.