Simplified Interface for B Estimation with Auto-Detection

A beginner-friendly wrapper for get_B that automatically detects data format, selects appropriate methods, and uses sensible defaults. This function minimizes the number of required parameters and handles most common use cases automatically.

get_B_auto(
  training_data,
  target_MAF,
  training_trait = NULL,
  target_trait = NULL,
  target_SE = NULL,
  target_case_prop = NULL,
  verbose = 1
)

Arguments

training_data

Either:

A data.frame with columns: MAF, BETA or (P/P_mlog10, N), and optionally rsID, CHR, POS. P_mlog10 represents -log10(p-value) and is preferred for extreme p-values.
An object of class glow_training_data from prepare_B_training_data

target_MAF

Numeric vector of target MAF values for which to predict effect sizes. Values should be in (0, 1)

training_trait

Character string specifying training trait type: "binary" or "continuous". If NULL and training_data is a glow_training_data object, automatically extracted. For data.frame input, this parameter is required

target_trait

Character string specifying target trait type: "binary" or "continuous". If NULL (default), uses training_trait

target_SE

Numeric value or vector of standard errors for target data. Required when target_trait == "continuous" and p-value method is used. Can be a single value (applied to all variants) or a vector matching length of target_MAF

target_case_prop

Numeric value or vector specifying the proportion of cases in target case-control study. Required when target_trait == "binary" and p-value method is used. This is the study case proportion (cases / total sample size), not population prevalence. Can be a single value or a vector matching length of target_MAF

verbose

Integer controlling verbosity: 0=silent, 1=warnings and summary (default), 2=detailed progress messages

Value

Numeric vector of predicted effect sizes (B) for each variant in target_MAF

Details

Automatic Features:

This function automatically:

Detects available data (BETA vs P-values) and chooses appropriate method
Extracts trait type from glow_training_data objects
Uses target_trait = training_trait if not specified
Selects best model using R-squared criterion
Detects both statistical and biological outliers (but keeps them)

Default Settings:

method = "auto" (use both methods if same trait + both data available, else pvalue)
selection_criterion = "R2" (maximize explained variance)
outlier_method = "both" (detect statistical + biological outliers)
outlier_action = "flag" (report but don't remove outliers)
return_full = FALSE (return simple vector)

When to Use:

Use this function when:

You're new to the GLOW framework
You want quick results with minimal setup
You trust automatic method selection
You don't need detailed diagnostics

For more control, use get_B_advanced or get_B directly.

Computational Complexity

O(8 * n_train + n_test) where n_train = number of training variants, n_test = number of target variants

Examples

if (FALSE) { # \dontrun{
# Example 1: Using glow_training_data object
training_data <- prepare_B_training_data(
  data = "gwas_results.csv",
  trait_type = "binary"
)

target_mafs <- seq(0.01, 0.4, by = 0.01)

B_estimates <- get_B_auto(
  training_data = training_data,
  target_MAF = target_mafs,
  target_case_prop = 0.5
)

# Example 2: Using data.frame directly
my_gwas <- data.frame(
  MAF = runif(100, 0.01, 0.5),
  BETA = rnorm(100, 0, 0.1),
  P = runif(100, 1e-8, 0.1),
  N = rep(5000, 100)
)

B_estimates <- get_B_auto(
  training_data = my_gwas,
  training_trait = "continuous",
  target_MAF = seq(0.01, 0.4, by = 0.01),
  target_SE = 0.05  # Single value for all variants
)

# Example 3: Cross-trait estimation (continuous to binary)
B_estimates <- get_B_auto(
  training_data = my_gwas,
  training_trait = "continuous",
  target_trait = "binary",
  target_MAF = seq(0.01, 0.4, by = 0.01),
  target_case_prop = 0.1
)
} # }