Add HDL shared test fixtures

README.md

@@ -397,7 +397,7 @@ The earth sciences folder contain subfolders for different data formats encounte
         - 'sk-n-as-severus-ont.svcf': a minimal SVCF file obtained from ONT data using severus and octopusv/correct (SK-N-AS cell line)
       - mills_and_1000G: Indels file downsampled based on reference position
       - savana
-        - 'sv_breakpoints.vcf': a minimal VCF file obtained from ONT data using SAVANA 
+        - 'sv_breakpoints.vcf': a minimal VCF file obtained from ONT data using SAVANA
       - vcfanno
         - 'vcfanno_grch38_module_test.tar.gz': exac.vcf.gz + exac.vcf.gz.tbi and they're reference ExAC vcf used to query
         - 'vcfanno.toml': configuration file for vcfanno to operate
@@ -741,6 +741,14 @@ The earth sciences folder contain subfolders for different data formats encounte
     - 1000GP.chr*.chunks.txt: chunks of the chromosome obtain with GLIMPSE_chunk
     - AFR.gwas: Study locus file. From [SuShiE](https://github.com/mancusolab/sushie).
     - AFR.ld: LD matrix file. From [SuShiE](https://github.com/mancusolab/sushie).
+    - hdl/reference/chr1.1_toy.bim: Synthetic toy HDL-format BIM sidecar for chunk 1.1, generated by `generate_toy_hdl_data.R` for HDL-compatible inputs.
+    - hdl/reference/chr1.1_toy.rda: Synthetic toy HDL-format LD reference payload for chunk 1.1, generated by `generate_toy_hdl_data.R` for HDL-compatible inputs.
+    - hdl/reference/chr1.2_toy.bim: Synthetic toy HDL-format BIM sidecar for chunk 1.2, generated by `generate_toy_hdl_data.R` for HDL-compatible inputs.
+    - hdl/reference/chr1.2_toy.rda: Synthetic toy HDL-format LD reference payload for chunk 1.2, generated by `generate_toy_hdl_data.R` for HDL-compatible inputs.
+    - hdl/reference/toy_snp_counter.RData: Synthetic toy HDL-format SNP count metadata, generated by `generate_toy_hdl_data.R` for HDL-compatible inputs.
+    - hdl/reference/toy_snp_list.RData: Synthetic toy HDL-format SNP list metadata, generated by `generate_toy_hdl_data.R` for HDL-compatible inputs.
+    - sumstats/trait1_canonical.tsv: Synthetic canonical toy summary statistics for trait 1, generated by `hdl/generate_toy_hdl_data.R` for small GWAS-style module inputs.
+    - sumstats/trait2_canonical.tsv: Synthetic canonical toy summary statistics for trait 2, generated by `hdl/generate_toy_hdl_data.R` for small GWAS-style module inputs.
   - popgen/clustering/: Test data for the new custom modules (`custom/clustering`, `custom/clustermetrics`, `custom/clustervisualization`) - `test.eigenvec`, `test_clusters.csv`, `test_features.tsv` + `README.md` (200-sample PCA derived from `popgen/plink_simulated` via `plink2 --pca`, plus KMeans cluster assignments)
   - svsig:
 

data/genomics/homo_sapiens/popgen/hdl/README.md

@@ -0,0 +1,41 @@
+# HDL Toy Test Dataset
+
+These files are synthetic toy fixtures for HDL module testing in the companion
+`nf-core/modules` work for `nf-core/modules#10912`. They are intended to exercise
+[HDL](https://github.com/zhenin/HDL) inputs in tests, not to provide a
+scientific LD reference panel or redistributed upstream reference bundle.
+
+## Layout
+
+- `reference/`: toy HDL LD reference chunks and metadata sidecars
+- `../sumstats/`: canonical toy summary-statistics tables aligned to the toy SNPs
+
+## Regeneration
+
+From this directory:
+
+```bash
+Rscript generate_toy_hdl_data.R
+```
+
+From the root of the `nf-core/test-datasets` worktree:
+
+```bash
+Rscript data/genomics/homo_sapiens/popgen/hdl/generate_toy_hdl_data.R
+```
+
+## R Objects
+
+The `.bim` sidecars, both canonical `sumstats/*.tsv` files, and the R binary
+payloads are all generated locally by `generate_toy_hdl_data.R` from fully
+synthetic constants in this directory.
+
+- `reference/chr1.1_toy.rda` and `reference/chr1.2_toy.rda` each contain
+  synthetic `LDsc`, `lam`, and `V` objects for one toy HDL chunk.
+- `reference/toy_snp_counter.RData` contains `nsnps.list` and
+  `nsnps.list.imputed`, each as a named one-element list with the toy chunk SNP
+  counts.
+- `reference/toy_snp_list.RData` contains `snps.list.imputed.vector`, the four
+  synthetic SNP IDs shared by the toy fixtures.
+- `../sumstats/trait1_canonical.tsv` and `../sumstats/trait2_canonical.tsv` are
+  tiny canonical summary-statistics tables keyed to those synthetic SNP IDs.

data/genomics/homo_sapiens/popgen/hdl/generate_toy_hdl_data.R

@@ -0,0 +1,111 @@
+#!/usr/bin/env Rscript
+# SPDX-License-Identifier: MIT
+
+args <- commandArgs(trailingOnly = FALSE)
+file_arg <- "--file="
+script_path <- sub(file_arg, "", args[grep(file_arg, args)])
+
+if (length(script_path) != 1 || script_path == "") {
+    stop("Unable to determine the script path from commandArgs().")
+}
+
+script_dir <- dirname(normalizePath(script_path))
+reference_dir <- file.path(script_dir, "reference")
+sumstats_dir <- file.path(script_dir, "..", "sumstats")
+
+dir.create(reference_dir, recursive = TRUE, showWarnings = FALSE)
+dir.create(sumstats_dir, recursive = TRUE, showWarnings = FALSE)
+
+writeLines(
+    c(
+        "1 rs1 0 101 A G",
+        "1 rs2 0 102 C T"
+    ),
+    file.path(reference_dir, "chr1.1_toy.bim")
+)
+
+writeLines(
+    c(
+        "1 rs3 0 201 A C",
+        "1 rs4 0 202 G T"
+    ),
+    file.path(reference_dir, "chr1.2_toy.bim")
+)
+
+lam <- c(1.3, 0.85)
+LDsc <- c(1.1, 1.4)
+V <- diag(2)
+save(
+    LDsc,
+    lam,
+    V,
+    file = file.path(reference_dir, "chr1.1_toy.rda"),
+    compress = "gzip"
+)
+
+lam <- c(1.25, 0.9)
+LDsc <- c(1.2, 1.35)
+V <- diag(2)
+save(
+    LDsc,
+    lam,
+    V,
+    file = file.path(reference_dir, "chr1.2_toy.rda"),
+    compress = "gzip"
+)
+
+nsnps.list <- list("1" = c(2, 2))
+nsnps.list.imputed <- list("1" = c(2, 2))
+save(
+    nsnps.list.imputed,
+    nsnps.list,
+    file = file.path(reference_dir, "toy_snp_counter.RData"),
+    compress = "gzip"
+)
+
+snps.list.imputed.vector <- c("rs1", "rs2", "rs3", "rs4")
+save(
+    snps.list.imputed.vector,
+    file = file.path(reference_dir, "toy_snp_list.RData"),
+    compress = "gzip"
+)
+
+trait1 <- data.frame(
+    SNP = c("rs1", "rs2", "rs3", "rs4"),
+    A1 = c("A", "C", "A", "G"),
+    A2 = c("G", "T", "C", "T"),
+    CHR = c(1, 1, 1, 1),
+    POS = c(101, 102, 201, 202),
+    RSID = c("rs1", "rs2", "rs3", "rs4"),
+    EffectAllele = c("A", "C", "A", "G"),
+    OtherAllele = c("G", "T", "C", "T"),
+    N = c(10000, 10000, 10000, 10000),
+    Z = c(0.5, -0.2, 0.4, -0.1)
+)
+write.table(
+    trait1,
+    file.path(sumstats_dir, "trait1_canonical.tsv"),
+    sep = "\t",
+    quote = FALSE,
+    row.names = FALSE
+)
+
+trait2 <- data.frame(
+    SNP = c("rs1", "rs2", "rs3", "rs4"),
+    A1 = c("A", "C", "A", "G"),
+    A2 = c("G", "T", "C", "T"),
+    CHR = c(1, 1, 1, 1),
+    POS = c(101, 102, 201, 202),
+    RSID = c("rs1", "rs2", "rs3", "rs4"),
+    EffectAllele = c("A", "C", "A", "G"),
+    OtherAllele = c("G", "T", "C", "T"),
+    N = c(12000, 12000, 12000, 12000),
+    Z = c(0.3, -0.4, 0.2, -0.2)
+)
+write.table(
+    trait2,
+    file.path(sumstats_dir, "trait2_canonical.tsv"),
+    sep = "\t",
+    quote = FALSE,
+    row.names = FALSE
+)

data/genomics/homo_sapiens/popgen/hdl/reference/chr1.1_toy.bim

@@ -0,0 +1,2 @@
+1 rs1 0 101 A G
+1 rs2 0 102 C T

data/genomics/homo_sapiens/popgen/hdl/reference/chr1.2_toy.bim

@@ -0,0 +1,2 @@
+1 rs3 0 201 A C
+1 rs4 0 202 G T

data/genomics/homo_sapiens/popgen/sumstats/README.md

@@ -0,0 +1,31 @@
+# Toy Population-Genetics Summary Statistics
+
+These files are tiny synthetic GWAS-style summary-statistics tables intended for
+module testing. They are generated from fixed constants by the companion HDL
+fixture generator at `../hdl/generate_toy_hdl_data.R`.
+
+## Layout
+
+- `trait1_canonical.tsv`: synthetic canonical summary statistics for trait 1
+- `trait2_canonical.tsv`: synthetic canonical summary statistics for trait 2
+
+## Regeneration
+
+From the `hdl/` directory:
+
+```bash
+Rscript generate_toy_hdl_data.R
+```
+
+From the root of the `nf-core/test-datasets` worktree:
+
+```bash
+Rscript data/genomics/homo_sapiens/popgen/hdl/generate_toy_hdl_data.R
+```
+
+## Notes
+
+These tables are not HDL-specific at the file-format level. They are kept under
+`popgen/sumstats/` so they can be reused by modules that consume small
+GWAS-style tabular inputs, while the HDL reference panel assets remain grouped
+under `popgen/hdl/reference/`.

data/genomics/homo_sapiens/popgen/sumstats/trait1_canonical.tsv

@@ -0,0 +1,5 @@
+SNP	A1	A2	CHR	POS	RSID	EffectAllele	OtherAllele	N	Z
+rs1	A	G	1	101	rs1	A	G	10000	0.5
+rs2	C	T	1	102	rs2	C	T	10000	-0.2
+rs3	A	C	1	201	rs3	A	C	10000	0.4
+rs4	G	T	1	202	rs4	G	T	10000	-0.1

data/genomics/homo_sapiens/popgen/sumstats/trait2_canonical.tsv

@@ -0,0 +1,5 @@
+SNP	A1	A2	CHR	POS	RSID	EffectAllele	OtherAllele	N	Z
+rs1	A	G	1	101	rs1	A	G	12000	0.3
+rs2	C	T	1	102	rs2	C	T	12000	-0.4
+rs3	A	C	1	201	rs3	A	C	12000	0.2
+rs4	G	T	1	202	rs4	G	T	12000	-0.2