In the above equation, z i is the z score indicating the strength of evidence, for example, differential representation score of a gene or a hairpin, in one source, say number i from total number of k sources: z i follows a standard normal distribution, so the integrated Z score also follows a standard Gaussian distribution assuming independence of all k evidences.
The combined two-tailed p value was calculated based on the integrated Z score and utilized such that p <0.05 significance cutoff corresponded to a minimum z comparative score of –1.96, the negative z score indicating a direction of depletion, positive indicating enrichment.
Metabolic risk was assessed using a composite z-score integrating standardized measurements of blood pressure, total cholesterol, high density lipoprotein, triglycerides and fat mass.
To integrate them into the same analysis, Z-score normalization3 was applied.
Availability: The Z-score calculation has been integrated in the QMEAN server available at: http://swissmodel.expasy.org/qmean.org/qmean
Mouse and human gene expression data were integrated and analyzed by hierarchical clustering using Z-score transformed expression values within each microarray dataset, the one minus un-centered correlation distance metric and complete linkage (GEO accession numbers: GSE25487 and GSE25488).
As can be seen, ComBat was successful in removing dataset-specific biases (Figure 2, Figure 3) since the samples of the merged data set integrated with ComBat were better intermixed than the samples merged with Z-score normalization.
Z-score analysis was used for intergroup evaluation.
The zeta- and z-score performance statistics are consequently modified to adapt the proposed method.
Z-scores were calculated as z-score = (x-mean)/sd.
Analyses were subsequently adjusted for other confounders, as defined below: smoking status, total cholesterol z-score, HDL z-score, creatinine z-score and BMI z-score (model 2).
