关于RobustRankAggreg R包AggregateRanks函数返回Scores（p值）是否已完成多重检验校正的技术咨询

Great question—this is a common point of confusion since the official docs don’t explicitly spell this out. Let me clarify:

• The p-values returned in the Scores column by AggregateRanks() are not adjusted for multiple testing. These are raw p-values calculated directly from the rank aggregation logic, with no built-in correction for the number of tests you’re evaluating.
• You will need to perform multiple testing correction on these scores yourself using standard R tools.

Example Workflow #

Here’s how to add adjusted p-values to your results after running AggregateRanks():

# Load the package
library(RobustRankAggreg)

# Sample rank data (replace with your own)
rank_datasets <- list(
  study1 = c("geneX", "geneY", "geneZ"),
  study2 = c("geneY", "geneX", "geneW"),
  study3 = c("geneZ", "geneX", "geneY")
)

# Run rank aggregation
aggregated_results <- AggregateRanks(rank_datasets, N = 4)

# Add FDR-adjusted p-values (choose method based on your needs)
aggregated_results$adjusted_p <- p.adjust(aggregated_results$Scores, method = "fdr")

# Inspect the updated results
print(aggregated_results)

Context for This Behavior #

Rank aggregation methods like the one implemented in AggregateRanks() prioritize combining rank data first, then generating raw p-values based on the aggregated distribution. Leaving correction to the user gives you flexibility to pick the method that aligns with your study design—whether that’s FDR, Bonferroni, Holm, or another approach.

内容的提问来源于stack exchange，提问作者durian