penetrance (version 0.1.0)

imputeAges: Impute Missing Ages in Family-Based Data

Description

Imputes missing ages in family-based data using a combination of Weibull distributions for affected individuals and empirical distributions for unaffected individuals. The function can perform both sex-specific and non-sex-specific imputations.

Usage

imputeAges(
  data,
  na_indices,
  baseline_male = NULL,
  baseline_female = NULL,
  alpha_male = NULL,
  beta_male = NULL,
  delta_male = NULL,
  alpha_female = NULL,
  beta_female = NULL,
  delta_female = NULL,
  baseline = NULL,
  alpha = NULL,
  beta = NULL,
  delta = NULL,
  max_age,
  sex_specific = TRUE,
  max_attempts = 100,
  geno_freq,
  trans,
  lik
)

Value

A data frame with the following modifications:

age

Updated with imputed ages for previously missing values

The rest of the data frame remains unchanged.

Arguments

data

A data frame containing family-based data with columns: family, individual, father, mother, sex, aff, age, geno, and isProband

na_indices

Vector of indices where ages need to be imputed

baseline_male, baseline_female

Data frames containing baseline age distributions for males/females

alpha_male, alpha_female

Shape parameters for male/female Weibull distributions

beta_male, beta_female

Scale parameters for male/female Weibull distributions

delta_male, delta_female

Location parameters for male/female Weibull distributions

baseline

Data frame containing overall baseline age distribution (non-sex-specific)

alpha, beta, delta

Overall Weibull parameters (non-sex-specific)

max_age

Maximum allowable age

sex_specific

Logical; whether to use sex-specific parameters

max_attempts

Maximum number of attempts for generating valid ages

geno_freq

Vector of genotype frequencies

trans

Transmission probabilities

lik

Likelihood matrix

Examples

# Create sample data with the same structure as used in mhChain
data <- data.frame(
  family = rep(1:2, each=5),
  individual = rep(1:5, 2),
  father = c(NA,1,1,1,1, NA,6,6,6,6),
  mother = c(NA,2,2,2,2, NA,7,7,7,7),
  sex = c(1,2,1,2,1, 1,2,1,2,1),
  aff = c(1,0,1,0,NA, 1,0,1,0,NA),
  age = c(45,NA,25,NA,20, 50,NA,30,NA,22),
  geno = c("1/2",NA,"1/2",NA,NA, "1/2",NA,"1/2",NA,NA),
  isProband = c(1,0,0,0,0, 1,0,0,0,0)
)

# Initialize parameters
na_indices <- which(is.na(data$age))
geno_freq <- c(0.999, 0.001)  # Frequency of normal and risk alleles
trans <- matrix(c(1,0,0.5,0.5), nrow=2)  # Transmission matrix
lik <- matrix(1, nrow=nrow(data), ncol=2)  # Likelihood matrix

# Create baseline data for both sex-specific and non-sex-specific cases
age_range <- 20:94
n_ages <- length(age_range)

# Sex-specific baseline data
baseline_male <- data.frame(
  age = age_range,
  cum_prob = (1:n_ages)/n_ages * 0.8  # Male cumulative probabilities
)

baseline_female <- data.frame(
  age = age_range,
  cum_prob = (1:n_ages)/n_ages * 0.9  # Female cumulative probabilities
)

# Non-sex-specific baseline data
baseline <- data.frame(
  age = age_range,
  cum_prob = (1:n_ages)/n_ages * 0.85  # Overall cumulative probabilities
)

# Example with sex-specific imputation
imputed_data_sex <- imputeAges(
  data = data,
  na_indices = na_indices,
  baseline_male = baseline_male,
  baseline_female = baseline_female,
  alpha_male = 3.5,
  beta_male = 20,
  delta_male = 20,
  alpha_female = 3.2,
  beta_female = 18,
  delta_female = 18,
  max_age = 94,
  sex_specific = TRUE,
  geno_freq = geno_freq,
  trans = trans,
  lik = lik
)

# Example with non-sex-specific imputation
imputed_data_nosex <- imputeAges(
  data = data,
  na_indices = na_indices,
  baseline = baseline,
  alpha = 3.3,
  beta = 19,
  delta = 19,
  max_age = 94,
  sex_specific = FALSE,
  geno_freq = geno_freq,
  trans = trans,
  lik = lik
)