Reconciles per-condition (univariate) SuSiE fine-mapping into a single set of
merged credible sets. Each row of the supplied
QtlFineMappingResult is treated as one condition (its
topLoci carrying that condition's credible sets); credible sets that
share variants across conditions are unioned via connected components, and
every variant is reported with its merged credible-set label plus the maximum
and median PIP across the conditions it appears in. A typical use is
selecting a representative lead variant per merged credible set to assemble
the "strong" input for mashPipeline.
Arguments
- fineMappingResult
A
QtlFineMappingResult(or anyFineMappingResult) produced by per-condition SuSiE fine-mapping. Each entry'stopLocimust carry a credible-set column (cs_<coverage*100>, e.g.cs_95, with values such as"susie_1"where the trailing integer is the set index and_0means "not in a credible set") and a PIP column.- coverage
Credible-set coverage level selecting the
cs_*column (default0.95->cs_95).
Value
A data.frame with one row per variant: variant_id,
credibleSetNames (the merged credible-set label), maxPip and
medianPip; or NULL when no credible sets are present.
Examples
data(qtlFineMappingExample)
mergeSusieCs(fineMappingResult = qtlFineMappingExample)
#> # A tibble: 14 × 4
#> variant_id credibleSetNames maxPip medianPip
#> <chr> <chr> <dbl> <dbl>
#> 1 chr22:32724254:T:C cs_1_1 0.0443 0.0443
#> 2 chr22:32724508:C:A cs_1_1 0.0443 0.0443
#> 3 chr22:32728910:G:A cs_1_1 0.0395 0.0395
#> 4 chr22:32729588:T:C cs_1_1 0.0345 0.0345
#> 5 chr22:32732510:G:A cs_1_1 0.0685 0.0685
#> 6 chr22:32734332:A:G cs_1_1 0.0663 0.0663
#> 7 chr22:32737583:C:T cs_1_1 0.0411 0.0411
#> 8 chr22:32737848:A:G cs_1_1 0.0663 0.0663
#> 9 chr22:32740783:T:C cs_1_1 0.0663 0.0663
#> 10 chr22:32743285:A:C cs_1_1 0.0663 0.0663
#> 11 chr22:32745710:T:C cs_1_1 0.0770 0.0770
#> 12 chr22:32745749:T:C cs_1_1 0.0364 0.0364
#> 13 chr22:32752021:A:G cs_1_1 0.0364 0.0364
#> 14 chr22:32755495:C:T cs_1_1 0.0365 0.0365