Skip to contents

Calculate Kinase Activity Scores using decoupleR

Source: R/kinase_activity.R
calcKinaseScore.Rd

calcKinaseScore calculates kinase activity scores based on input data and a specified network of regulatory relationships (decoupler network).

Usage

calcKinaseScore(
  resTab,
  decoupler_network,
  corrThreshold = 0.9,
  statType = c("stat", "log2FC"),
  nPerm = 100
)

Arguments

resTab

A data frame containing the input data with columns site, stat, and log2FC.

decoupler_network

A data frame representing the decoupleR network with columns source and target.

corrThreshold

A numeric value specifying the correlation threshold for filtering correlated regulons. Default is 0.9.

statType

A character string specifying the type of statistic to use. Options are "stat" or "log2FC". Default is "stat".

nPerm

A numeric value specifying the number of permutations for the null distribution. Default is 100.

Value

A data frame with kinase activity scores, including columns for `source`, `score`, and `p_value`.

Details

The function performs the following steps:

  1. Removes duplicate rows based on the site column.

  2. Filters the data to include only those sites present in the target column of the decoupler network.

  3. Prepares the input table based on the specified statType.

  4. Intersects the input table with the decoupler network to find common regulons.

  5. Checks for correlated regulons and filters out those exceeding the correlation threshold.

  6. Calculates kinase activity using a weighted mean approach.

  7. Processes the results to handle NA values and formats the output.

Examples

resTab <- data.frame(
site = c("EGFR_Y1172", "EGFR_Y1197", "EGFR_S1166", "ROCK2_S1374",
"WASL_Y256", "GAB1_Y259", "ADD1_S586", "EPHA2_Y772", "PRKDC_T2638",
"PRKDC_T2609", "PRKDC_S2612"),
stat = c(-10.038770, -5.945562, 5.773384, -7.303834, 5.585326, 5.971104,
5.199119, -5.169500, 5.130228, 5.407387, 4.493933),
log2FC = c(-2.6113343, -2.4858615, 1.0056629, -1.1561780, 1.6421145,
2.0296634, 1.3766283, -0.8531656, 1.0742881, 1.0042942, 1.0608129)
)
decoupler_network <- data.frame(
source = c(rep("ABL1", 5), rep("CDK2", 6)),
mor = c(rep(1, 11)),
target = c("EGFR_Y1172", "EGFR_Y1197", "EGFR_S1166", "ROCK2_S1374",
"WASL_Y256", "GAB1_Y259", "ADD1_S586", "EPHA2_Y772", "PRKDC_T2638",
"PRKDC_T2609", "PRKDC_S2612"),
likelihood = c(rep(1, 11))
)
# Call the function
calcKinaseScore(resTab, decoupler_network)
#> # A tibble: 2 × 3
#>   source score p_value
#>   <chr>  <dbl>   <dbl>
#> 1 ABL1   -2.39    0.72
#> 2 CDK2    3.51    0.18