get_robust_colocalization: Recalibrate and summarize robust colocalization events.

In colocboost: Multi-Context Colocalization Analysis for QTL and GWAS Studies

Recalibrate and summarize robust colocalization events.

Description

get_robust_colocalization get the colocalization by discarding the weaker colocalization events or colocalized outcomes

Usage

get_robust_colocalization(
  cb_output,
  cos_npc_cutoff = 0.5,
  npc_outcome_cutoff = 0.2,
  pvalue_cutoff = NULL,
  weight_fudge_factor = 1.5,
  coverage = 0.95
)

Arguments

cb_output	Output object from colocboost analysis
cos_npc_cutoff	Minimum threshold of normalized probability of colocalization (NPC) for CoS.
npc_outcome_cutoff	Minimum threshold of normalized probability of colocalized traits in each CoS.
pvalue_cutoff	Maximum threshold of marginal p-values of colocalized variants on colocalized traits in each CoS.
weight_fudge_factor	The strength to integrate weight from different outcomes, default is 1.5
coverage	A number between 0 and 1 specifying the “coverage” of the estimated colocalization confidence sets (CoS) (default is 0.95).

Value

A "colocboost" object with some or all of the following elements:

cos_summary	A summary table for colocalization events.
vcp	The variable colocalized probability for each variable.
cos_details	A object with all information for colocalization results.
data_info	A object with detailed information from input data
model_info	A object with detailed information for colocboost model
ucos_from_cos	A object with information for trait-specific effects if exists after removing weaker signals.

Source

See detailed instructions in our tutorial portal: https://statfungen.github.io/colocboost/articles/Interpret_ColocBoost_Output.html

See Also

Other colocboost_inference: get_ambiguous_colocalization(), get_colocboost_summary()

Examples

# colocboost example
set.seed(1)
N <- 1000
P <- 100
# Generate X with LD structure
sigma <- 0.9^abs(outer(1:P, 1:P, "-"))
X <- MASS::mvrnorm(N, rep(0, P), sigma)
colnames(X) <- paste0("SNP", 1:P)
L <- 3
true_beta <- matrix(0, P, L)
true_beta[10, 1] <- 0.5 # SNP10 affects trait 1
true_beta[10, 2] <- 0.4 # SNP10 also affects trait 2 (colocalized)
true_beta[50, 2] <- 0.3 # SNP50 only affects trait 2
true_beta[80, 3] <- 0.6 # SNP80 only affects trait 3
Y <- matrix(0, N, L)
for (l in 1:L) {
  Y[, l] <- X %*% true_beta[, l] + rnorm(N, 0, 1)
}
res <- colocboost(X = X, Y = Y)
res$cos_details$cos$cos_index
filter_res <- get_robust_colocalization(res, cos_npc_cutoff = 0.5, npc_outcome_cutoff = 0.2)
filter_res$cos_details$cos$cos_index

colocboost documentation built on June 8, 2025, 11:07 a.m.