Package 'DANDELION'

Organize DANDELION Gene-Gene Results

Description

Cleans significant DANDELION pairs when the exposure side contains genes and merges nearby exposure genes into genomic loci.

Usage

calc_pair.gene(
  mat.sig,
  mat.p,
  p.wes,
  gene1,
  ref.table.keep,
  eta.wgs = 1e-05,
  verbose = FALSE
)

Arguments

mat.sig	Matrix encoding significant pairs, returned by med_gene().
mat.p	Matrix of DANDELION p-values, returned by med_gene().
p.wes	Named numeric vector of gene-level trait association p-values.
gene1	Candidate exposure genes returned by med_gene().
ref.table.keep	Gene annotation data frame after filtering. Required columns are gene_name, Chromosome, start, and end.
eta.wgs	Significance threshold for WES genes. Default is 1e-5.
verbose	Logical. If TRUE, progress messages are printed. Default is FALSE.

Value

A list containing pairs_dact, gene.pair, sig_gene2, and non_sig.gene2.

---

Organize DANDELION SNP-Gene Results

Description

Cleans significant DANDELION pairs when the exposure side contains SNPs. SNPs are mapped to cis genes using uniq_snp first. If a SNP is not present in uniq_snp, the function optionally maps it to the nearest or overlapping gene using SNP.ref and ref.table.keep.

Usage

calc_pair.snp(
  mat.sig,
  mat.p,
  p.wes,
  gene1,
  uniq_snp,
  ref.table.keep,
  eta.wgs = 1e-05,
  SNP.ref = NULL,
  verbose = FALSE
)

Arguments

mat.sig	Matrix encoding significant pairs, returned by med_gene().
mat.p	Matrix of DANDELION p-values, returned by med_gene().
p.wes	Named numeric vector of gene-level trait association p-values.
gene1	Candidate SNPs returned by med_gene().
uniq_snp	Data frame containing known SNP-to-cis-gene mapping. Required columns are SNP and GeneSymbol.
ref.table.keep	Gene annotation data frame after filtering. Required columns are gene_name, Chromosome, start, and end.
eta.wgs	Significance threshold for WES genes. Default is 1e-5.
SNP.ref	Optional SNP annotation data frame with columns SNP, SNPPos, and SNPChr. Used only for SNPs not mapped by uniq_snp.
verbose	Logical. If TRUE, progress messages are printed. Default is FALSE.

Value

A list containing pairs_dact, gene.pair, sig_gene2, and non_sig.gene2.

---

Generate DANDELION Network Figure

Description

Generates a network plot from DANDELION gene-pair results.

Usage

gen_fig(gene.pair, p.wes, eta.wgs = 1e-05, pic_dir)

Arguments

gene.pair	Data frame of identified gene pairs, returned by calc_pair.gene() or calc_pair.snp(). Must contain columns region and gene2.
p.wes	Named numeric vector of gene-level trait association p-values.
eta.wgs	Significance threshold for WES genes. Default is 1e-5.
pic_dir	Directory to save the figure.

Value

Invisibly returns the path to the saved PDF file.

---

Apply DANDELION to Identify Candidate Disease Proximal Genes

Description

Applies the DANDELION procedure to integrate trans-regulatory association p-values and gene-level trait association p-values. The exposure side can be either distal genes or SNPs. For clarity, exposures are referred to as gene1, and candidate disease proximal genes are referred to as gene2.

Usage

med_gene(
  p.trans,
  p.wes,
  ref.table,
  gene1.list,
  target.fdr = 0.1,
  dist = 5e+06,
  gene1.type = c("SNP", "Gene"),
  SNP.ref = NULL,
  n.cores = 1,
  verbose = FALSE
)

Arguments

p.trans	A numeric matrix of trans-regulatory association p-values. Rows are candidate disease proximal genes (gene2), and columns are exposures (gene1), either distal genes or SNPs. Row names and column names must be provided. (e.g., from trans-eQTL studies like eQTLGen)
p.wes	A named numeric vector of gene-level trait association p-values (e.g., from WES burden tests or GWAS-based gene-level tests). Note: Ensure that names (Gene Symbols) match the row names of p.trans
ref.table	A data frame containing gene annotation and genomic positions. Required columns are gene_name, type, Chromosome, start, and end. Chromosome should use the format chr1, chr2, etc.
gene1.list	A character vector of candidate exposures to analyze. These values must be a subset of colnames(p.trans).
target.fdr	False Discovery Rate threshold. DANDELION uses this to decide which gene pairs are statistically significant. Default is 0.1.
dist	The cis-window size (in base pairs). Genes within this distance from the exposure will be excluded to focus on true distal (trans) effects. Default is 5e6.
gene1.type	Exposure type. Must be either "SNP" or "Gene".
SNP.ref	Optional SNP annotation data frame used when gene1.type = "SNP". Required columns are SNP, SNPPos, and SNPChr. SNPChr can be 1 or chr1; both formats are accepted.
n.cores	Number of cores for parallel execution. On non-Windows systems, values larger than 1 use parallel::mclapply(). On Windows, the function falls back to single-core execution.
verbose	Logical. If TRUE, progress messages are printed. Default is FALSE.

Value

A list with three elements: gene1, the analyzed exposures with at least one valid DANDELION result; mat.sig, a matrix encoding significant gene1-gene2 pairs; and mat.p, a matrix of DANDELION p-values.

Examples

set.seed(1)
p.trans <- matrix(runif(60, 0.001, 0.9), nrow = 12, ncol = 5)
rownames(p.trans) <- paste0("G", 1:12)
colnames(p.trans) <- paste0("E", 1:5)
p.wes <- runif(12, 0.001, 0.9)
names(p.wes) <- paste0("G", 1:12)
ref.table <- data.frame(
  gene_name = c(paste0("G", 1:12), paste0("E", 1:5)),
  type = "protein_coding",
  Chromosome = "chr1",
  start = seq_len(17) * 1e7,
  end = seq_len(17) * 1e7 + 1000
)
res <- med_gene(
  p.trans = p.trans,
  p.wes = p.wes,
  ref.table = ref.table,
  gene1.list = colnames(p.trans),
  gene1.type = "Gene"
)

---

Print DANDELION pair results

Description

Print DANDELION pair results

Usage

## S3 method for class 'dandelion_pairs'
print(x, ...)

Arguments

x	A dandelion_pairs object returned by calc_pair.gene() or calc_pair.snp().
...	Additional arguments, currently unused.

Value

Invisibly returns x.

---

Print a DANDELION result object

Description

Print a DANDELION result object

Usage

## S3 method for class 'dandelion_result'
print(x, ...)

Arguments

x	A dandelion_result object returned by med_gene().
...	Additional arguments, currently unused.

Value

Invisibly returns x.

---