Fine-mapping QTLs and GWAS with the pecotmr SuSiE Wrapper

Overview

Fine-mapping reduces a region with one or more associations to a small set of credible causal variants. pecotmr wraps SuSiE (Sum of Single Effects) and SuSiE-RSS (its summary-statistics analog) and adds:

A two-stage fit — SuSiE-inf first, then SuSiE initialized from the SuSiE-inf result — that improves credible-set purity in regions with background polygenic signal.
Unified post-processing producing the same 22-column top_loci table regardless of whether the input was individual-level data or summary statistics, and regardless of whether you fit SuSiE, SuSiE-inf, SuSiE-RSS, fSuSiE, or mvSuSiE.
Pipeline-level entry points that bundle loading, QC, fine-mapping, and (for individual-level data) TWAS-weight learning.

This vignette demonstrates all three workflows:

Data you have	Entry point	Example dataset
Individual-level genotype + phenotype (QTL studies)	`univariate_analysis_pipeline()`	`eqtl_region_example`
QTL summary statistics + LD reference	`rss_analysis_pipeline()`	placeholder (real QTL nominal output)
GWAS summary statistics + LD reference	`rss_analysis_pipeline()`	placeholder (real GWAS sumstats)

For SuSiE theory, see the susieR vignettes; this vignette focuses on how to drive it through pecotmr.

library(pecotmr)

Quick start: individual-level fine-mapping

eqtl_region_example contains a genotype matrix X (415 samples × 2,828 variants) and a residualized molecular phenotype y_res. Below we run the full pipeline — initial PIP-screening, SuSiE-inf, SuSiE initialised from SuSiE-inf, post-processing, and (optionally) TWAS-weight learning — in a single call.

data(eqtl_region_example)
X <- eqtl_region_example$X
y <- as.numeric(eqtl_region_example$y_res)
maf <- apply(X, 2, function(g) {
  f <- mean(g, na.rm = TRUE) / 2
  min(f, 1 - f)
})

# Subset to a smaller window for a fast vignette build; the same call works
# on the full region.
set.seed(42)
sub <- sort(sample(ncol(X), 500))

fit <- univariate_analysis_pipeline(
  X = X[, sub], Y = y, maf = maf[sub],
  region = "chr22:32119000-33000000",
  twas_weights = FALSE,
  finemapping_extra_opts = list(refine = FALSE),
  verbose = 0
)

The return value is a list with the fitted models and the unified top_loci table:

names(fit)

## [1] "susie_inf_fitted"         "susie_fitted"            
## [3] "finemapping_result"       "sumstats"                
## [5] "other_quantities"         "top_loci"                
## [7] "susie_inf_result_trimmed" "total_time_elapsed"

dim(fit$top_loci)

## [1] 22 22

head(fit$top_loci[, c("variant", "pip", "cs_95", "cs_70",
                       "cs_95_purity", "method")], 5)

##              variant        pip   cs_95   cs_70 cs_95_purity method
## 1 chr22:32526384:T:G 0.02736065 susie_0 susie_0            0  susie
## 2 chr22:32538691:A:G 0.02736065 susie_0 susie_0            0  susie
## 3 chr22:32595330:A:G 0.12089319 susie_0 susie_0            0  susie
## 4 chr22:32607002:T:C 0.03388883 susie_0 susie_0            0  susie
## 5 chr22:32607795:T:C 0.03324444 susie_0 susie_0            0  susie

The methods column shows which model each row came from — the pipeline runs both SuSiE-inf and SuSiE by default and row-binds their top loci. For most downstream consumers, the SuSiE rows are what you want; SuSiE-inf rows are useful as a sanity check on inferred non-sparse effects.

Choosing among SuSiE variants

pecotmr exposes five SuSiE-family fine-mapping variants. Pick which to fit via the methods parameter on univariate_analysis_pipeline() (individual-level) or rss_analysis_pipeline() / susie_rss_pipeline() (summary statistics):

Variant	Individual-level	Summary statistics	What it models
SuSiE alone	`methods = "susie"`	`methods = "susie_rss"`	Sparse causal effects only. No infinitesimal/background term. Lightest weight; can fold polygenic background into noisy “extra” credible sets.
SuSiE with SuSiE-inf init (default)	`methods = c("susie_inf", "susie")` + `add_susie_inf = TRUE`	`methods = c("susie_inf_rss", "susie_rss")` + `add_susie_inf = TRUE`	Two-stage: SuSiE-inf absorbs background polygenicity first, SuSiE then refines sparse signals from that initialisation. Recommended default for most QTL and GWAS fine-mapping.
SuSiE-inf alone	`methods = "susie_inf"`	`methods = "susie_inf_rss"`	Sparse + infinitesimal joint fit. Useful for regions with strong polygenic background where the SuSiE refinement step is undesired.
SuSiE-ash alone	`methods = "susie_ash"`	`methods = "susie_ash_rss"`	Sparse + adaptive shrinkage (ASH) mixture. Useful when you suspect a mix of medium-effect and small-effect variants beyond a small handful of large ones.
SuSiE-ash with SuSiE-inf init	`methods = c("susie_inf", "susie_ash")` + `add_susie_inf = TRUE`	`methods = c("susie_inf_rss", "susie_ash_rss")` + `add_susie_inf = TRUE`	Same chained two-stage idea as the SuSiE variant but with ASH on the downstream fit: SuSiE-inf absorbs polygenic background first, then SuSiE-ash refines using the mixture prior.

The chained SuSiE-inf → SuSiE two-stage is the default both for backward compatibility and because it produces clean credible sets in most settings: SuSiE alone can fold polygenic background into noisy “extra” credible sets; SuSiE-inf absorbs that background first so the SuSiE refinement concentrates on sparse signals.

add_susie_inf = TRUE chains SuSiE-inf into any of "susie" or "susie_ash" that appear in methods. Listing all three (methods = c("susie_inf", "susie", "susie_ash") with add_susie_inf = TRUE) fits SuSiE-inf once, then both SuSiE and SuSiE-ash initialised from it — useful for direct head-to-head comparison on the same warm start.

All five populate the same 22-column top_loci table. When you fit multiple methods in one call, the rows are row-bound and labelled by the method column.

Note: SuSiE-inf chain init only carries information when the SuSiE-inf fit finds non-zero mappable effects. On a region where SuSiE-inf explains everything via its infinitesimal component (all V == 0, all mu == 0), upstream susieR correctly treats the init as empty, and the chained downstream fit collapses to the independent run. This is the right behaviour, not a bug — but it means head-to-head differences only show up on regions with at least one sparse signal.

Backward compatibility with `add_susie_inf`

The pre-existing add_susie_inf boolean still works exactly as before when methods is NULL (the default):

add_susie_inf = TRUE (default) → fit SuSiE-inf, then chain it into SuSiE. Equivalent to methods = c("susie_inf", "susie").
add_susie_inf = FALSE → plain SuSiE alone. Equivalent to methods = "susie".

When methods is set explicitly, add_susie_inf controls whether the chained-init shortcut is applied to any "susie" or "susie_ash" method that appears alongside "susie_inf". Set add_susie_inf = FALSE to fit each requested method independently rather than chained.

Fitting multiple variants in one call

multi <- univariate_analysis_pipeline(
  X = X[, sub], Y = y, maf = maf[sub],
  methods = c("susie_inf", "susie", "susie_ash"),
  twas_weights = FALSE,
  signal_cutoff = 0
)
# top_loci has rows for all three methods; filter on `method` to inspect each
table(multi$top_loci$method)

Lower-level entry points

If you want the raw fits without pecotmr’s post-processing:

# Two-stage convenience for the recommended default:
fits <- fit_susie_inf_then_susie(X[, sub], y)         # list(susie, susie_inf)
fits_rss <- fit_susie_inf_then_susie_rss(z, R, n)

# Or call susieR directly with the right unmappable_effects flag:
inf_fit <- susieR::susie(X, y, unmappable_effects = "inf",
                          convergence_method = "pip", refine = FALSE)
ash_fit <- susieR::susie(X, y, unmappable_effects = "ash",
                          convergence_method = "pip")

# Then post-process any combination through the unified pipeline:
post <- postprocess_finemapping_fits(
  fits = list(susie_inf = inf_fit, susie_ash = ash_fit),
  data_x = X[, sub], data_y = y, maf = maf[sub],
  coverage = 0.95, secondary_coverage = c(0.7, 0.5)
)
out <- format_finemapping_output(post, primary_method = "susie_inf")

The unified `top_loci` table

postprocess_finemapping_fits() produces a single table with the following 22 columns, in this order:

#	Column	Source
1	`#chr`	parsed from `variant`
2	`start`	`pos - 1` (BED 0-based)
3	`end`	`pos` (BED half-open)
4	`a1`	from parsed `variant`
5	`a2`	from parsed `variant`
6	`variant`	the row’s variant id (`chr:pos:A2:A1`)
7	`gene`	phenotype identifier (`colnames(data_y)[1]`)
8	`event`	`paste(condition_id, gene, sep = "_")`
9	`n`	analyzed sample count
10	`maf`	from the `maf` argument
11	`beta`	univariate effect (`sumstats$betahat`)
12	`se`	univariate SE (`sumstats$sebetahat`)
13	`pip`	per-variant SuSiE PIP
14	`posterior_effect_mean`	`colSums(alpha * mu)`
15	`posterior_effect_se`	`sqrt(pmax(colSums(alpha * mu2) - mean^2, 0))`
16	`cs_95`	CS membership at coverage 0.95 (see below)
17	`cs_70`	CS membership at coverage 0.70
18	`cs_50`	CS membership at coverage 0.50
19	`cs_95_purity`	0.95-coverage CS purity for this row
20	`method`	`"susie"`, `"susie_inf"`, `"susie_rss"`, …
21	`grange_start`	parsed from the `region` argument
22	`grange_end`	parsed from the `region` argument

CS membership strings

cs_95, cs_70, cs_50 are character strings, not integers. Each value is <method>_<cs_index> where <cs_index> is the per-method 1-based credible-set number:

susie_1, susie_2, … for SuSiE credible sets
susie_inf_1, … for SuSiE-inf credible sets
<method>_0 is the sentinel for “PIP-only retention, not in any CS at this coverage”

CS indices are not sequenced across methods — each method numbers its credible sets from 1.

Row uniqueness

The row identity is (variant, gene, method, cs_membership):

A variant in CS 1 and CS 2 of the same method (overlapping CS) gives two rows.
A variant in both susie CS 1 and susie_inf CS 1 gives two rows with the same (variant, gene) but different method.
A PIP-retained variant not in any CS gives one row per method with cs_* = "<method>_0" and cs_95_purity = 0.

This shape preserves overlapping-CS evidence both within and across methods. Downstream consumers (xqtl-protocol, transMap, ColocBoost, TWAS) read the columns above; the schema is the contract.

QTL summary-statistics fine-mapping

When you have z-scores instead of X and Y, use rss_analysis_pipeline(). It bundles:

load_rss_data() — tabix-region-extracts the sumstats and standardises column names via a YAML mapping (see the RSS QC vignette for the YAML schema).
summary_stats_qc() — runs allele harmonization, SLALOM (default) or DENTIST QC, and optional RAISS imputation. The vignette referenced above covers QC in detail.
susie_rss_pipeline() — fits the two-stage SuSiE-RSS model and post-processes into the unified top_loci table.

Pipeline call (file-based, production shape)

# TODO(data): replace with real QTL sumstats and LD-metadata paths
qtl_sumstat_path  <- "<path/to/qtl_nominal.tsv.gz>"
qtl_column_map    <- "<path/to/qtl_column_map.yml>"
ld_meta_path      <- "<path/to/ld_meta.tsv>"
region            <- "<chr:start-end>"
gene_id           <- "<ENSG...>"
n_qtl_sample      <- 0  # 0 = read from sumstats

ld_data <- load_LD_matrix(ld_meta_path, region = region)

fm <- rss_analysis_pipeline(
  sumstat_path        = qtl_sumstat_path,
  column_file_path    = qtl_column_map,
  LD_data             = ld_data,
  n_sample            = n_qtl_sample,
  region              = region,
  extract_region_name = gene_id,
  region_name_col     = 4L,
  qc_method           = "slalom",
  finemapping_method  = "susie_rss",
  finemapping_opts    = list(L = 20, L_greedy = 5,
                             coverage = c(0.95, 0.7, 0.5),
                             signal_cutoff = 0.025,
                             min_abs_corr = 0.8),
  impute              = TRUE
)
head(fm$rss_data_analyzed[[1]]$top_loci)

Lower-level alternative (in-memory, runnable here)

For demonstration on the shipped example data, we skip the file-based loading and use susie_rss_pipeline() directly with the sumstats and LD already in memory. This is what rss_analysis_pipeline() calls internally after QC.

data(gwas_sumstats_example)
data(eqtl_region_example)

# We treat the shipped GWAS-style sumstats as if they were QTL z-scores
# for this demonstration. In a real workflow you would load QTL nominal
# output via load_rss_data().
sumstats <- gwas_sumstats_example
X_ref    <- eqtl_region_example$X
R        <- compute_LD(X_ref, method = "sample")
R_sub    <- R[sumstats$variant_id, sumstats$variant_id]

fm_rss <- susie_rss_pipeline(
  sumstats          = sumstats,
  LD_mat            = R_sub,
  n                 = nrow(X_ref),
  L                 = 10, L_greedy = 5,
  analysis_method   = "susie_rss",
  coverage          = 0.95,
  secondary_coverage = c(0.7, 0.5)
)
names(fm_rss)

## [1] "finemapping_result" "sumstats"           "top_loci"

dim(fm_rss$top_loci)

## [1] 20 22

head(fm_rss$top_loci[, c("variant", "pip", "cs_95",
                          "cs_95_purity", "method")], 5)

##              variant        pip       cs_95 cs_95_purity    method
## 1 chr22:32337604:C:T 0.19820547 susie_rss_3    0.9846607 susie_rss
## 2 chr22:32350609:A:G 0.09322469 susie_rss_3    0.9846607 susie_rss
## 3 chr22:32354393:A:G 0.14380650 susie_rss_3    0.9846607 susie_rss
## 4 chr22:32362470:T:C 0.37141549 susie_rss_3    0.9846607 susie_rss
## 5 chr22:32362793:A:G 0.14992585 susie_rss_3    0.9846607 susie_rss

Note that the output schema is the same 22 columns as the individual-level pipeline — method is "susie_rss" instead of "susie", but everything downstream (ColocBoost, xqtl-enrichment, TWAS) consumes the same table.

GWAS summary-statistics fine-mapping

GWAS fine-mapping uses the same rss_analysis_pipeline() / susie_rss_pipeline() call. The only differences are at the loading layer:

Column mapping. GWAS conventions (effect/non-effect, OR vs beta, per-variant n) differ from QTL nominal output. The RSS QC vignette shows the YAML schema; common GWAS alternatives are effect_allele/other_allele for A1/A2, OR (you’ll need log(OR) for beta), and N for n_sample.
No extract_region_name. GWAS sumstats are not gene-keyed, so you slice by physical region only. Pass region = "chr:start-end" and omit extract_region_name / region_name_col.
Per-variant sample size. GWAS sumstats often have a per-variant n_sample column (some variants were imputed in fewer cohorts); load_rss_data() reads that automatically when the column is mapped.

# TODO(data): replace with real GWAS sumstats and LD-metadata paths
gwas_sumstat_path <- "<path/to/gwas.tsv.gz>"
gwas_column_map   <- "<path/to/gwas_column_map.yml>"
ld_meta_path      <- "<path/to/ld_meta.tsv>"
region            <- "<chr:start-end>"

ld_data <- load_LD_matrix(ld_meta_path, region = region)

fm_gwas <- rss_analysis_pipeline(
  sumstat_path       = gwas_sumstat_path,
  column_file_path   = gwas_column_map,
  LD_data            = ld_data,
  n_case             = 0,        # set if known; 0 reads from sumstats
  n_control          = 0,
  region             = region,
  qc_method          = "slalom",
  finemapping_method = "susie_rss",
  impute             = TRUE
)

The shipped gwas_finemapping_example dataset shows what the output looks like for a real GWAS region:

data(gwas_finemapping_example)
str(gwas_finemapping_example, max.level = 2)

## List of 1
##  $ :List of 4
##   ..$ alpha: num [1:10, 1:2828] 6.65e-29 0.00 1.08e-11 0.00 0.00 ...
##   .. ..- attr(*, "dimnames")=List of 2
##   ..$ pip  : Named num [1:2828] 1.08e-11 1.08e-11 3.36e-11 1.04e-11 1.05e-11 ...
##   .. ..- attr(*, "names")= chr [1:2828] "chr22:32119788:T:C" "chr22:32119867:T:G" "chr22:32119961:T:G" "chr22:32120053:T:C" ...
##   ..$ V    : num [1:10] 3.16e-04 1.28e+01 1.19e-04 1.29e+01 1.29e+01 ...
##   ..$ sets :List of 5

length(gwas_finemapping_example[[1]]$pip)

## [1] 2828

sum(gwas_finemapping_example[[1]]$pip > 0.1)

## [1] 10

gwas_finemapping_example ships in the legacy susie_result_trimmed shape used by older pecotmr pipelines (PIPs, alpha matrix, credible sets). New workflows should consume the 22-column top_loci table produced by rss_analysis_pipeline() instead.

Downstream consumers

The top_loci table feeds directly into several other pecotmr workflows:

Enrichment (xqtl_enrichment_wrapper): computes prior enrichment of QTL fine-mapping signals at GWAS PIPs. See the enrichment vignette.
Colocalization (colocboost_pipeline): uses the trimmed SuSiE fits — finemapping_result slot — as colocalization input. See the ColocBoost vignette.
TWAS (twas_weights_pipeline, twas_pipeline): uses the SuSiE effect estimates as weight learners for transcriptome-wide association testing. See the TWAS weights vignette and the TWAS Z-score vignette.

Summary

Function	Use it when
`univariate_analysis_pipeline()`	Individual-level data: you have `X` (samples × variants), `Y` (phenotype). Pick variant(s) via `methods = c("susie", "susie_inf", "susie_ash")`.
`rss_analysis_pipeline()`	Summary statistics (QTL or GWAS): you have z-scores in a tabix file and an LD reference. Pick variant(s) via `methods = c("susie_rss", "susie_inf_rss", "susie_ash_rss")`.
`susie_rss_pipeline()`	Lower-level: you already have in-memory `z`, `R` (or `X`), `n`. Same `methods` argument as `rss_analysis_pipeline()`.
`fit_susie_inf_then_susie()` / `fit_susie_inf_then_susie_rss()`	Lowest-level convenience for the two-stage chained default (SuSiE-inf init → SuSiE).
`postprocess_finemapping_fits()` + `format_finemapping_output()`	Post-process any combination of `susie`, `susie_inf`, `susie_ash`, `susie_rss`, `susie_inf_rss`, `susie_ash_rss`, `mvsusie`, `fsusie` fits into the unified 22-column `top_loci`.

All five entry points produce — or can be post-processed into — the same 22-column top_loci table documented above, which is the contract all downstream pecotmr consumers read.

sessionInfo()

## R version 4.5.3 (2026-03-11)
## Platform: x86_64-conda-linux-gnu
## Running under: Ubuntu 24.04.4 LTS
## 
## Matrix products: default
## BLAS/LAPACK: /home/runner/work/pecotmr/pecotmr/.pixi/envs/default/lib/libopenblasp-r0.3.33.so;  LAPACK version 3.12.0
## 
## locale:
##  [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8       
##  [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8   
##  [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C          
## [10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C   
## 
## time zone: Etc/UTC
## tzcode source: system (glibc)
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] pecotmr_0.5.3
## 
## loaded via a namespace (and not attached):
##   [1] DBI_1.3.0                   coloc_5.2.3                
##   [3] bitops_1.0-9                gridExtra_2.3              
##   [5] rlang_1.2.0                 magrittr_2.0.5             
##   [7] otel_0.2.0                  matrixStats_1.5.0          
##   [9] susieR_0.16.4               compiler_4.5.3             
##  [11] RSQLite_3.53.1              GenomicFeatures_1.62.0     
##  [13] colocboost_1.0.9            png_0.1-9                  
##  [15] systemfonts_1.3.2           vctrs_0.7.3                
##  [17] quadprog_1.5-8              stringr_1.6.0              
##  [19] pkgconfig_2.0.3             crayon_1.5.3               
##  [21] fastmap_1.2.0               XVector_0.50.0             
##  [23] Rsamtools_2.26.0            rmarkdown_2.31             
##  [25] tzdb_0.5.0                  UCSC.utils_1.6.1           
##  [27] ragg_1.5.2                  purrr_1.2.2                
##  [29] bit_4.6.0                   xfun_0.57                  
##  [31] Rfast_2.1.5.2               cachem_1.1.0               
##  [33] cigarillo_1.0.0             GenomeInfoDb_1.46.2        
##  [35] jsonlite_2.0.0              blob_1.3.0                 
##  [37] reshape_0.8.10              DelayedArray_0.36.0        
##  [39] tictoc_1.2.1                BiocParallel_1.44.0        
##  [41] irlba_2.3.7                 parallel_4.5.3             
##  [43] R6_2.6.1                    VariantAnnotation_1.56.0   
##  [45] bslib_0.11.0                stringi_1.8.7              
##  [47] RColorBrewer_1.1-3          rtracklayer_1.70.1         
##  [49] GenomicRanges_1.62.1        jquerylib_0.1.4            
##  [51] Rcpp_1.1.1-1.1              Seqinfo_1.0.0              
##  [53] SummarizedExperiment_1.40.0 knitr_1.51                 
##  [55] R.utils_2.13.0              readr_2.2.0                
##  [57] IRanges_2.44.0              Matrix_1.7-5               
##  [59] tidyselect_1.2.1            viridis_0.6.5              
##  [61] abind_1.4-8                 yaml_2.3.12                
##  [63] codetools_0.2-20            curl_7.1.0                 
##  [65] plyr_1.8.9                  lattice_0.22-9             
##  [67] tibble_3.3.1                Biobase_2.70.0             
##  [69] KEGGREST_1.50.0             S7_0.2.2                   
##  [71] evaluate_1.0.5              desc_1.4.3                 
##  [73] RcppParallel_5.1.11-2       Biostrings_2.78.0          
##  [75] pillar_1.11.1               MatrixGenerics_1.22.0      
##  [77] stats4_4.5.3                ieugwasr_1.1.0             
##  [79] generics_0.1.4              vroom_1.7.1                
##  [81] RCurl_1.98-1.19             hms_1.1.4                  
##  [83] S4Vectors_0.48.0            ggplot2_4.0.3              
##  [85] scales_1.4.0                glue_1.8.1                 
##  [87] tools_4.5.3                 MungeSumstats_1.18.1       
##  [89] BiocIO_1.20.0               data.table_1.17.8          
##  [91] BSgenome_1.78.0             GenomicAlignments_1.46.0   
##  [93] fs_2.1.0                    XML_3.99-0.22              
##  [95] grid_4.5.3                  tidyr_1.3.2                
##  [97] AnnotationDbi_1.72.0        restfulr_0.0.16            
##  [99] cli_3.6.6                   zigg_0.0.2                 
## [101] textshaping_1.0.5           viridisLite_0.4.3          
## [103] mixsqp_0.3-54               S4Arrays_1.10.1            
## [105] dplyr_1.2.1                 gtable_0.3.6               
## [107] R.methodsS3_1.8.2           sass_0.4.10                
## [109] digest_0.6.39               BiocGenerics_0.56.0        
## [111] SparseArray_1.10.8          rjson_0.2.23               
## [113] htmlwidgets_1.6.4           farver_2.1.2               
## [115] memoise_2.0.1               htmltools_0.5.9            
## [117] pkgdown_2.2.0               R.oo_1.27.1                
## [119] lifecycle_1.0.5             httr_1.4.8                 
## [121] bit64_4.8.2                 MASS_7.3-65

pecotmr authors

2026-06-11

Overview

Quick start: individual-level fine-mapping

Choosing among SuSiE variants

Backward compatibility with `add_susie_inf`

Fitting multiple variants in one call

Lower-level entry points

The unified `top_loci` table

CS membership strings

Row uniqueness

QTL summary-statistics fine-mapping

Pipeline call (file-based, production shape)

Lower-level alternative (in-memory, runnable here)

GWAS summary-statistics fine-mapping

Downstream consumers

Summary

Fine-mapping QTLs and GWAS with the pecotmr SuSiE Wrapper

pecotmr authors

2026-06-11

Overview

Quick start: individual-level fine-mapping

Choosing among SuSiE variants

Backward compatibility with add_susie_inf

Fitting multiple variants in one call

Lower-level entry points

The unified top_loci table

CS membership strings

Row uniqueness

QTL summary-statistics fine-mapping

Pipeline call (file-based, production shape)

Lower-level alternative (in-memory, runnable here)

GWAS summary-statistics fine-mapping

Downstream consumers

Summary

Backward compatibility with `add_susie_inf`

The unified `top_loci` table