Gene expression and metabolism preceding soft scald, a chilling injury of ‘Honeycrisp’ apple fruit

BMC Genomics

Table 2 Summary of data analyses procedure for both transcriptomic and metabolomic data of ‘Honeycrisp’ apple fruit to predict postharvest soft scald risk

Data type	Timepoints	Objective	Analysis
Transcriptomic, normalized	0–8 wk	Latent trends	Principal components analysis (PCA)
normalized	0–8 wk	Differential expression in the multivariate case	ANOVA-simultaneous component analysis
raw counts	0 wk	Pre-harvest environmental effects	RobiNA edge R differential expression according to soft scald risk^a
raw counts	2 wk	Effects of chilling	RobiNA edge R differential expression according to soft scald risk^a
	-	Functional characterization of significant transcripts	Gene Ontology overrepresentation analysis
Metabolomic, means centered, standard deviation squared	0–8 wk	Latent trends	Principal components analysis (PCA)
	0–8 wk	Differential expression in the multivariate case	ANOVA-simultaneous component analysis
	0 wk	Pre-harvest environmental effects	t-test/ Pearson’s correlation + Ward clustering algorithm heatmap visualization
	2 wk	Effects of chilling	t-test/ Pearson’s correlation + Ward clustering algorithm heatmap visualization
Transcriptomic and metabolomic, normalized and set to a common scale by means centering	0–8 wk	Correlation among expressed genes and metabolites	Pearson’s correlation via Cytoscape Expression Correlation plug-in at r = │0.95│, followed by network subtraction with the Network Analyzer plug-in

Data sets were analyzed separately to find differences between high and low risk orchards at-harvest, after 2 weeks cold storage, and overall
^aSoft scald risk assigned categorically as “low risk”/“high risk” based on disorder incidence at 12 weeks of storage. “Low risk”, no fruit exhibited soft scald symptoms. “High risk”, the presence of any fruit with soft scald symptoms

ISSN: 1471-2164