The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
This vignette demonstrates how to perform single-trait fine-mapping
analysis using FineBoost, a specialized single-trait version of
ColocBoost, with both individual-level data and summary statistics.
Specifically focusing on the 2nd trait with 2 causal variants (194 and
589) from the Ind_5traits
and Sumstat_5traits
datasets included in the package.
In this section, we demonstrate how to perform fine-mapping using
individual-level genotype (X
) and phenotype
(Y
) data. This approach uses raw data directly to identify
causal variants.
# Load example data
data(Ind_5traits)
X <- Ind_5traits$X[[2]]
Y <- Ind_5traits$Y[[2]]
res <- colocboost(X = X, Y = Y)
#> Validating input data.
#> Starting gradient boosting algorithm.
#> Gradient boosting for outcome 1 converged after 44 iterations!
#> Performing inference on colocalization events.
colocboost_plot(res)
This section demonstrates fine-mapping analysis using summary statistics along with a proper LD matrix.
# Load example data
data(Sumstat_5traits)
sumstat <- Sumstat_5traits$sumstat[[2]]
LD <- get_cormat(Ind_5traits$X[[2]])
res <- colocboost(sumstat = sumstat, LD = LD)
#> Validating input data.
#> Starting gradient boosting algorithm.
#> Gradient boosting for outcome 1 converged after 44 iterations!
#> Performing inference on colocalization events.
colocboost_plot(res)
In scenarios where LD information is unavailable, FineBoost can still perform fine-mapping under the assumption that there is a single causal variant. This approach is less computationally intensive but assumes that only one variant within a region is causal.
# Load example data
res <- colocboost(sumstat = sumstat)
#> Validating input data.
#> Warning in colocboost(sumstat = sumstat): Providing the LD for summary
#> statistics data is highly recommended. Without LD, only a single iteration will
#> be performed under the assumption of one causal variable per outcome.
#> Additionally, the purity of CoS cannot be evaluated!
#> Starting gradient boosting algorithm.
#> Running ColocBoost with assumption of one causal per outcome per region!
#> Performing inference on colocalization events.
colocboost_plot(res)
Note: Weak learners SEL in FineBoost may capture
noise as putative signals, potentially introducing false positives to
our findings. To identify and filter spurious signals, we discard
fine-tunned the threshold of \(\Delta
L_l\) using extensive simulations to balance sensitivity and
specificity. This threshold is set to 0.025 by default for ColocBoost
when detect the colocalization, but we suggested a less conservative
threshold of 0.015 for FineBoost when performing single-trait
fine-mapping analysis (check_null_max = 0.015
as we
suggested).
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.