This vignette demonstrates how to perform multi-trait colocalization
analysis using a mixed-type dataset, including both individual level
data and summary statistics. ColocBoost provides a flexible framework to
integrate data both at the individual level or at the summary statistic
level, allowing the handling of scenarios where the individual data is
available for some traits (like xQTLs) and the summary data is available
for other traits (disease/trait GWAS).
1. Loading individual and summary statistics data
To get started, load both Ind_5traits and
Sumstat_5traits datasets into your R session. Once loaded,
create a mixed dataset as follows:
- For traits 1, 2, 3, 4: use individual level genotype and phenotype
data.
- For trait 5: use summary statistics data.
- Note that
LD could be calculated from the
X data in the Ind_5traits dataset, but it is
not included in the Sumstat_5traits dataset.
Causal variant structure
The dataset features two causal variants with indices 194 and
589.
- Causal variant 194 is associated with traits 1, 2, 3, and 4.
- Causal variant 589 is associated with traits 2, 3, and 5 (summary
level data).
# Load example data
data(Ind_5traits)
data(Sumstat_5traits)
# Create a mixed dataset
X <- Ind_5traits$X[1:4]
Y <- Ind_5traits$Y[1:4]
sumstat <- Sumstat_5traits$sumstat[5]
LD <- get_cormat(Ind_5traits$X[[1]])
For analyze a specific one type of data, you can refer to the
following tutorials Individual
Level Data Colocalization and Summary
Level Data Colocalization.
Due to the file size limitation of CRAN release, this is a subset of
simulated data. See full dataset in colocboost paper
repo.
2. ColocBoost in disease-agnostic mode
The preferred format for colocalization analysis in ColocBoost using
mixed-type dataset:
- Individual level data:
X and
Y are organized as lists, matched by trait index,
(X[1], Y[1]) contains individual level data for trait
1,(X[2], Y[2]) contains individual level data for trait
2,- And so on for each trait under analysis.
- Summary level data:
sumstat is organized as a list of data.frames for all
traitsLD is a matrix of linkage disequilibrium (LD)
information for all variants across all traits.
This function requires specifying genotypes X and
phenotypes Y from the individual-level dataset and summary
statistics sumstat and LD matrix LD from
summary dataset:
# Run colocboost
res <- colocboost(X = X, Y = Y, sumstat = sumstat, LD = LD)
#> Validating input data.
#> Starting gradient boosting algorithm.
#> Gradient boosting for outcome 4 converged after 40 iterations!
#> Gradient boosting for outcome 5 converged after 59 iterations!
#> Gradient boosting for outcome 1 converged after 61 iterations!
#> Gradient boosting for outcome 3 converged after 91 iterations!
#> Gradient boosting for outcome 2 converged after 94 iterations!
#> Performing inference on colocalization events.
# Identified CoS
res$cos_details$cos$cos_index
#> $`cos1:y1_y2_y3_y4`
#> [1] 186 194 168 205
#>
#> $`cos2:y2_y3_y5`
#> [1] 589 593
3. ColocBoost in disease-prioritized mode
When integrating with GWAS data, to mitigate the risk of biasing
early updates toward stronger xQTL signals in LD proximity-and should in
fact colocalize with-weaker signals from the disease trait of interest,
we also implement a disease-prioritized mode of ColocBoost as a soft
prioritization strategy to colocalize variants with putative causal
effects on the focal trait.
To run the disease-prioritized mode, you need to specify the
focal_outcome_idx argument in the colocboost()
function.
focal_outcome_idx indicates the index of the focal
trait in all traits of interest.- Note that the
focal_outcome_idx are counted from
individual level data to summary statistics.
Therefore, in this example, focal_outcome_idx = 5 is used
to indicate the index of the focal trait from (4 individual level
traits) and (1 summary statistics).
# Run colocboost
res <- colocboost(X = X, Y = Y,
sumstat = sumstat, LD = LD,
focal_outcome_idx = 5)
#> Validating input data.
#> Starting gradient boosting algorithm.
#> Gradient boosting for focal outcome 5 converged after 29 iterations!
#> Gradient boosting for outcome 4 converged after 60 iterations!
#> Gradient boosting for outcome 1 converged after 82 iterations!
#> Gradient boosting for outcome 3 converged after 97 iterations!
#> Gradient boosting for outcome 2 converged after 99 iterations!
#> Performing inference on colocalization events.
# Plotting the focal only results colocalization results
colocboost_plot(res, plot_focal_only = TRUE)
Unlike the other existing methods, the disease-prioritized mode of
ColocBoost only used a soft prioritization strategy. Therefore, it not
only identify the colocalization of the focal trait, but also the
colocalization across the other traits without the focal trait. To
extract all CoS and visualization of all colocalization results, you can
use the following code:
# Identified CoS
res$cos_details$cos$cos_index
#> $`cos1:y1_y2_y3_y4`
#> [1] 186 194 168 205
#>
#> $`cos2:y2_y3_y5:merged`
#> [1] 589 593
# Plotting all results
colocboost_plot(res)
