Compute fold-change or convert between log-ratio and fold-change.

foldchange computes the fold change for two sets of values. logratio2foldchange converts values from log-ratios to fold changes. foldchange2logratio does the reverse.

foldchange(num, denom)

logratio2foldchange(logratio, base = 2)

foldchange2logratio(foldchange, base = 2)

Arguments

num, denom

vector/matrix of numeric values

logratio

vector/matrix of log-ratio values

base

Exponential base for the log-ratio.

foldchange

vector/matrix of fold-change values

Value

A vector or matrix of the same dimensions as the input containing the converted values.

Details

Fold changes are commonly used in the biological sciences as a mechanism for comparing the relative size of two measurements. They are computed as: \(\frac{num}{denom}\) if \(num>denom\), and as \(\frac{-denom}{num}\) otherwise.

Fold-changes have the advantage of ease of interpretation and symmetry about \(num=denom\), but suffer from a discontinuity between -1 and 1, which can cause significant problems when performing data analysis. Consequently statisticians prefer to use log-ratios.

Functions

• foldchange(): Compute fold-change.

• logratio2foldchange(): Compute foldchange from log-ratio values.

• foldchange2logratio(): Compute log-ratio from fold-change values.

Author

Gregory R. Warnes greg@warnes.net

Examples

a <- 1:21
b <- 21:1

f <- foldchange(a, b)

cbind(a, b, f)
#>        a  b          f
#>  [1,]  1 21 -21.000000
#>  [2,]  2 20 -10.000000
#>  [3,]  3 19  -6.333333
#>  [4,]  4 18  -4.500000
#>  [5,]  5 17  -3.400000
#>  [6,]  6 16  -2.666667
#>  [7,]  7 15  -2.142857
#>  [8,]  8 14  -1.750000
#>  [9,]  9 13  -1.444444
#> [10,] 10 12  -1.200000
#> [11,] 11 11   1.000000
#> [12,] 12 10   1.200000
#> [13,] 13  9   1.444444
#> [14,] 14  8   1.750000
#> [15,] 15  7   2.142857
#> [16,] 16  6   2.666667
#> [17,] 17  5   3.400000
#> [18,] 18  4   4.500000
#> [19,] 19  3   6.333333
#> [20,] 20  2  10.000000
#> [21,] 21  1  21.000000