The coefficient of similarity (F) values between PFGE profiles were calculated as described [ 5].
Distances between individual profiles were calculated using Euclidean, while Ward's was used to determine the clustering.
Background correlation coefficients between expression profiles were calculated for each randomly shuffled version of the selected subset, and the mean value and variance of these coefficients were calculated for the 1,000 shuffled subsets.
The distances between expression profiles were calculated based on Pearson correlation, and the average-linkage method from the Eisen Cluster program was used to cluster genes and tumor samples.
Pearson correlations between gene expression profiles were calculated, and agglomerative hierarchical clustering revealed the existence of three large clusters of gene interactions.
In the study of Kim et al., similarities between gene expression profiles were calculated using the Pearson correlation coefficient, and gene expression profile similarities (distances) were converted into two dimensions using force-directed placement.
Correlations between BGL before neuropsychological assessment and IQ scores and between BGL before scanning and metabolite profiles were calculated and ranged from r = 0.001 to r = −0.242 (all P >0.05).
To determine correlation between the microbial community composition and the metabolic profiles, distance matrices using Manhattan distances [16] for microbial and metabolic profiles were calculated independently, then Pearson correlation coefficients between the two distance matrices were calculated.
Relative-to-neutral profiles were calculated as point-by-point differences between either velocity or acceleration profiles of non-neutral conditions (happy, disgust, anger) and the neutral profile presented in the same block (relevant/irrelevant).
First, a matrix of distances between expression profiles is calculated using a distance or similarity measure, such as Pearson's centered correlation coefficient (PCC).
