Overview: SPACox implements an empirical saddlepoint approximation (SPA) for GWAS of time‑to‑event traits, providing accurate p‑values for low‑frequency and rare variants. SPAmix extends SPACox to handle population structure, while WtCoxG further accounts for sample relatedness and case ascertainment and leverages external allele frequencies to increase statistical power.

Features of the Methods:

Method Population structure Sample relatedness Case ascertainment
SPACox Not Not Not
SPAmix Modeled Not Not
WtCoxG Not Modeled Modeled

SPACox

SPACox is the baseline method for analyzing unrelated subjects in a homogeneous population.

Citations:

Bi et al. (2020). Fast and accurate method for genome-wide time-to-event data analysis and its application to UK Biobank. American Journal of Human Genetics. doi:10.1016/j.ajhg.2020.06.003

Step 1: Model Fitting and Preprocessing

A quick example is provided below. Refer to ?GRAB.NullModel and ?GRAB.SPACox for detailed parameter instructions.

# Load data
PhenoFile = system.file("extdata", "simuPHENO.txt", package = "GRAB")
PhenoData = data.table::fread(PhenoFile, header = T)

# Step 1, time-to-event trait, SPACox
obj.SPACox = GRAB.NullModel(
  survival::Surv(SurvTime, SurvEvent) ~ AGE + GENDER, 
  data = PhenoData, 
  subjIDcol = "IID", 
  method = "SPACox", 
  traitType = "time-to-event"
)

Step 2: Association Testing

Refer to ?GRAB.Marker and ?GRAB.SPACox for detailed parameter instructions.

# Step 2, SPACox
GenoFile = system.file("extdata", "simuPLINK.bed", package = "GRAB")
OutputFile = file.path(tempdir(), "Results_SPACox.txt")

# Marker-level testing
GRAB.Marker(obj.SPACox, GenoFile = GenoFile, OutputFile = OutputFile)

# Read results
head(data.table::fread(OutputFile))

Output Columns:

  • Marker: Variant identifier
  • Info: CHR:POS:REF:ALT
  • AltFreq: Alternative allele frequency
  • AltCounts: Alternative allele count
  • MissingRate: Proportion missing
  • Pvalue: Association p-value
  • zScore: Test statistic

SPAmix

SPAmix extends SPACox to support complex population structures, including admixed ancestry and multiple populations. It does not account for sample relatedness.

Citations:

Ma et al. (2025). Sparse estimation of high-dimensional genetic correlation and its application to global biobank meta-analysis. Genome Biology. doi:10.1186/s13059-025-03827-9

Step 1: Model Fitting and Preprocessing

PC_columns in the control list is a comma-separated column names of SNP-derived principal components (e.g., "PC1,PC2") and is required by SPAmix. A quick example is provided below. Refer to ?GRAB.NullModel and ?GRAB.SPAmix for detailed parameter instructions.

# Load data
PhenoFile = system.file("extdata", "simuPHENO.txt", package = "GRAB")
PhenoData = data.table::fread(PhenoFile, header = T)

# Step 1, time-to-event trait, SPAmix
obj.SPAmix = GRAB.NullModel(
  Surv(SurvTime, SurvEvent) ~ AGE + GENDER + PC1 + PC2, 
  data = PhenoData, 
  subjIDcol = "IID", 
  method = "SPAmix", 
  traitType = "time-to-event", 
  control = list(PC_columns = "PC1,PC2")
)

Step 2: Association Testing

Refer to ?GRAB.Marker and ?GRAB.SPAmix for detailed parameter instructions.

# Step 2, SPAmix
GenoFile = system.file("extdata", "simuPLINK.bed", package = "GRAB")
OutputFile = file.path(tempdir(), "Results_SPAmix.txt")

# Marker-level testing
GRAB.Marker(obj.SPAmix, GenoFile = GenoFile, OutputFile = OutputFile)

# Read results
head(data.table::fread(OutputFile))

Output Columns:

  • Pheno: Phenotype identifier (pheno_1, pheno_2, …)
  • Marker: Variant identifier
  • Info: CHR:POS:REF:ALT
  • AltFreq: Alternative allele frequency
  • AltCounts: Alternative allele count
  • MissingRate: Proportion missing
  • Pvalue: Association p-value
  • zScore: Test statistic

WtCoxG

WtCoxG is a Cox-based association test for time‑to‑event traits that accounts for sample relatedness and corrects for case ascertainment.

Key Features:

  • Corrects for case ascertainment bias in biobank studies
  • Leverages external MAF from public resources (e.g., gnomAD)
  • Incorporates sample relatedness via sparse GRM
  • Performs batch effect QC between study cohort and reference population
  • Saddlepoint approximation (SPA) provides accurate p-values, especially for rare variants and extreme case-control ratios

Citation:

Li et al. (2025). High-powered, robust, and versatile survival analysis via weighted Cox regression. Nature Computational Science. doi:10.1038/s43588-025-00864-z

Step 1: Model Fitting and Preprocessing

A quick example is provided below. Refer to ?GRAB.NullModel and ?GRAB.WtCoxG for detailed parameter instructions.

# Load files
PhenoFile <- system.file("extdata", "simuPHENO.txt", package = "GRAB")
PhenoData <- data.table::fread(PhenoFile, header = TRUE)
SparseGRMFile <- system.file("extdata", "SparseGRM.txt", package = "GRAB")
GenoFile <- system.file("extdata", "simuPLINK.bed", package = "GRAB")
RefAfFile <- system.file("extdata", "simuRefAf.txt", package = "GRAB")
OutputFile <- file.path(tempdir(), "resultWtCoxG.txt")

# Fit null model and test batch effects
obj.WtCoxG <- GRAB.NullModel(
  survival::Surv(SurvTime, SurvEvent) ~ AGE + GENDER,
  data = PhenoData,
  subjIDcol = "IID",
  method = "WtCoxG",
  traitType = "time-to-event",
  GenoFile = GenoFile,
  SparseGRMFile = SparseGRMFile,
  RefAfFile = RefAfFile,
  RefPrevalence = 0.1
)

WtCoxG specific mandatory parameters

  • RefAfFile: Reference allele frequency file (see format below)
  • RefPrevalence: Population disease prevalence (0 < p < 0.5)
  • obj.WtCoxG contains the data structure for step 2

RefAfFile Format

The reference allele frequency file is whitespace-delimited with the following columns:

CHROM   POS      ID          REF  ALT  AF_ref   AN_ref
1       10177    rs367896724  A    C   0.4258   251390
1       10235    rs540538026  T    A   0.0009   251306
1       10352    rs555500075  T    A   0.4104   251480

Format specifications:

  • CHROM: Chromosome
  • POS: Position
  • ID: Variant identifier
  • REF: Reference allele
  • ALT: Alternative allele
  • AF_ref: Allele frequency in reference population
  • AN_ref: Allele number in reference population

Step 2: Association Testing

Refer to ?GRAB.Marker and ?GRAB.WtCoxG for detailed parameter instructions.

# Marker-level testing
GRAB.Marker(obj.WtCoxG, GenoFile, OutputFile)

# View results
head(data.table::fread(OutputFile))

Output Columns:

  • WtCoxG.ext: P-value using external MAF
  • WtCoxG.noext: P-value without external MAF
  • Marker: Variant identifier
  • Info: CHR:POS:REF:ALT
  • AltFreq: Alternative allele frequency in study
  • AltCounts: Alternative allele count
  • MissingRate: Proportion of missing genotypes
  • AF_ref: Reference allele frequency
  • AN_ref: Reference allele number
  • pvalue_bat: Batch effect test p-value
  • TPR: True positive rate estimate
  • sigma2: Variance parameter estimate
  • w.ext: Optimal external weight
  • var.ratio.w0: Variance ratio for internal analysis
  • var.ratio.int: Variance ratio adjustment
  • var.ratio.ext: Variance ratio with external MAF
This site uses Just the Docs, a documentation theme for Jekyll.