Methods for Penetrance Estimation in Family-Based Studies

Implements statistical methods for estimating disease penetrance in family-based studies. Penetrance refers to the probability of disease manifestation in individuals carrying specific genetic variants. The package provides tools for age-specific penetrance estimation, handling missing data, and accounting for ascertainment bias in family studies. Cite as: Kubista, N., Braun, D. & Parmigiani, G. (2025) .

Penetrance R package

An R package for the estimation of age-specific penetrance for complex family-based studies in a format compatible with the Fam3PRO R package.

Motivation

Accurate estimation of age-specific penetrance is essential for assessing disease risk in individuals with pathogenic genetic variants (PGVs). Penetrance refers to the probability that an individual carrying a specific genetic variant will develop the associated trait, such as cancer. Estimating this probability is a crucial step in clinical decision-making and personalized risk assessment for hereditary cancer syndromes.

The package leverages Mendelian inheritance models, which are widely used in family-based genetic studies to assess how genetic variants are passed down through generations. These models typically involve a proband—an individual for whom family history and genetic data are collected. The proband serves as the starting point for mapping out the family’s genetic structure, including relationships and phenotypic traits, such as cancer diagnoses. Family data, including cancer occurrence, ages of diagnosis, and genetic test results, are collected for the proband and their relatives. Using these data, Mendelian models compute the likelihood of genetic configurations and disease outcomes based on inheritance patterns.

The core methodology in the package relies on a four-parameter Weibull distribution to model age-specific penetrance. Estimation is performed using a Bayesian framework with Markov chain Monte Carlo (MCMC) methods, allowing the package to provide robust and flexible penetrance estimates. Through this approach, the package models the likelihood of cancer occurrence across family members, even when some genotypic information is missing or incomplete, as is common in real-world studies.

The package also incorporates prior knowledge into the estimation process, enabling users to specify default, custom, or study-based prior distributions. By employing the Elston–Stewart peeling algorithm, the package efficiently calculates likelihoods across family pedigrees, ensuring scalability and accuracy, even in large datasets.

By providing functions for data input, prior specification, and estimation, the package equips researchers and clinicians with tools for estimating cancer risk in complex family-based studies. This supports informed decision-making and preventive strategies in hereditary cancer syndromes, where understanding the genetic basis of risk is critical for patient care.

Citation

If you use the penetrance package, please cite:

Nicolas Kubista, Danielle Braun, Giovanni Parmigiani (2025). The penetrance R package for estimation of age specific risk in family-based studies. Bioinformatics Advances, Volume 5, Issue 1, vbaf154. https://doi.org/10.1093/bioadv/vbaf154

Installation

To install, use

git clone [email protected]:bayesmendel/penetrance.git

Open the source directory as a new R project and install the package with

devtools::install()

or install it directly from GitHub in RStudio:

devtools::install_github("bayesmendel/penetrance")

Quick-start guide

The following is a quick-start guide for basic usage of the package. For greater detail on available options, please refer to the other articles.

The primary function in the package is penetrance(). The package workflow includes three main parts: supplying family data and specifying the estimation settings, estimating the posterior distribution using MCMC, and returning samples from the approximated posterior distribution representing the estimated penetrance function.

library(penetrance)

Pedigree

The user must specify the pedigree argument as a list of data frames, where each data frame contains one family’s data. Each data frame must have the following columns:

• PedigreeID: A numeric or character identifier for the family. It must be consistent for all members of the family.

• ID: A numeric or character identifier for each individual. It must be unique within a pedigree.

• Sex: An integer where 0 indicates female and 1 indicates male. Unknown sex can be coded as NA.

• MotherID: The ID of the individual’s mother, or NA if the mother is not included in the pedigree.

• FatherID: The ID of the individual’s father, or NA if the father is not included in the pedigree.

• isProband: An integer where 1 indicates that the individual is a proband and 0 otherwise.

• CurAge: An integer indicating the age of censoring: current age if the individual is alive or age at death if deceased. Values must be between 1 and max_age; unknown ages can be coded as NA.

• isAff: An integer indicating affection status for the cancer of interest, with 1 for diagnosed individuals and 0 for unaffected individuals. Unknown status can be coded as NA.

• Age: An integer indicating age at cancer diagnosis. It should be NA if isAff is 0 or NA; otherwise it must be between 1 and max_age and should not exceed CurAge.

• Geno: An integer indicating germline genetic test results, with 1 for carriers and 0 for non-carriers. Unknown or untested individuals can be coded as NA.

Model specification

Available options include:

• pedigree: A list of data frames containing pedigrees in the format described above.
• twins: A list specifying identical twins or triplets. For example, list(c("ora024", "ora027"), c("aey063", "aey064")) specifies two identical-twin pairs.
• n_chains: Integer, the number of chains for parallel computation. Default is 1.
• n_iter_per_chain: Integer, the number of iterations for each chain. Default is 10000.
• ncores: Integer, the number of cores for parallel computation. Default is 6.
• baseline_data: Absolute age-specific baseline risk data. With sex_specific = TRUE, this is a data frame with Male and Female columns; with sex_specific = FALSE, it is a numeric vector or single-column data frame.
• max_age: Integer, the maximum age considered for analysis. Default is 94.
• remove_proband: Logical, indicating whether to remove probands from the analysis. Default is FALSE.
• age_imputation: Logical, indicating whether to perform age imputation. Default is FALSE.
• median_max: Logical, indicating whether to use the baseline median age or max_age as an upper bound for the median proposal. Default is TRUE.
• BaselineNC: Logical, indicating that non-carrier penetrance is assumed to equal baseline penetrance. Only TRUE is currently supported.
• var: Numeric vector containing proposal variances for the Metropolis–Hastings algorithm.
• burn_in: Numeric, the fraction of results to discard as burn-in. Default is 0.
• thinning_factor: Integer, the factor by which to thin the results. Default is 1.
• imp_interval: Integer, the interval at which age imputation is performed when age_imputation = TRUE. Default is 100.
• distribution_data: Data used to generate prior distributions.
• allele_freq: Numeric, the population allele frequency of the risk variant. Default is 0.0001.
• sample_size: Optional numeric sample size for distribution generation.
• ratio: Optional numeric ratio for distribution generation.
• prior_params: A list containing prior-distribution parameters.
• risk_proportion: Numeric, the risk proportion used for distribution generation.
• summary_stats: Logical, indicating whether to include summary statistics in the output. Default is TRUE.
• rejection_rates: Logical, indicating whether to include rejection rates in the output. Default is TRUE.
• density_plots: Logical, indicating whether to include density plots in the output. Default is TRUE.
• plot_trace: Logical, indicating whether to include trace plots in the output. Default is TRUE.
• penetrance_plot: Logical, indicating whether to include cumulative penetrance plots in the output. Default is TRUE.
• penetrance_plot_pdf: Logical, indicating whether to include penetrance density plots in the output. Default is TRUE.
• plot_loglikelihood: Logical, indicating whether to include log-likelihood plots in the output. Default is TRUE.
• plot_acf: Logical, indicating whether to include autocorrelation plots for posterior samples. Default is TRUE.
• probCI: Numeric, the probability level for credible intervals in penetrance plots. Default is 0.95.
• sex_specific: Logical, indicating whether to use sex-specific parameters in the analysis. Default is TRUE.

Additional user inputs

• The penetrance() function takes absolute age-specific probabilities of developing cancer as the baseline_data input. With the default BaselineNC = TRUE, this baseline is assumed to represent non-carrier penetrance. For rare variants, this is considered a reasonable assumption.

• The allele frequency (allele_freq) defaults to 0.0001. The function converts it to carrier prevalence using the approximation carrier_prevalence ≈ 2 × allele_freq for rare autosomal dominant conditions.

• The penetrance() function includes an option for automatic age imputation through age_imputation. Ages are imputed during the MCMC routine based on affection status, sex, and degree of relationship to the proband who carries the pathogenic variant.

• Monozygotic twins or triplets can be specified using the twins argument. For example:

twins <- list(c("ora024", "ora027"), c("aey063", "aey064"))