Semi-Parametric Survival Estimation for pedigrees

Flora Alarcon,Gregory Nuel,Violaine Plante-Bordeneuve

DOI: https://doi.org/10.48550/arXiv.1607.04215

2016-07-15

Abstract:Mendelian diseases are determined by a single mutation in a given gene. However, in the case of diseases with late onset, the age at onset is variable; it can even be the case that the onset is not observed in a lifetime. Estimating the survival function of the mutation carriers and the effect of modifying factors such as the sex, mutation, origin, etc, is a task of importance, both for management of mutation carriers and for prevention. In this work, we present a semi-parametric method based on a proportional to estimate the survival function using pedigrees ascertained through affected individuals (probands). Not all members of the pedigree need to be genotyped. The ascertainment bias is corrected by using only the phenotypic information from the relatives of the proband, and not of the proband himself. The method manage ungenotyped individuals through belief propagation in Bayesian networks and uses an EM algorithm to compute a Kaplan-Meier estimator of the survival function. The method is illustrated on simulated data and on a samples of families with transthyretin-related hereditary amyloidosis, a rare autosomal dominant disease with highly variable age of onset.

Applications

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to accurately estimate the survival function of individuals carrying mutations and the factors affecting the age of onset in families with late - onset Mendelian diseases. Specifically, the paper focuses on how to estimate the survival function through family - based data (especially families identified through affected individuals) and correct for ascertainment bias in the case of incomplete genotype information. Ascertainment bias refers to the estimation bias that may be caused by the way in which research subjects are selected (for example, selecting families through affected individuals). The paper proposes a semi - parametric method, which combines the Cox proportional hazards model and the belief propagation algorithm in Bayesian networks, and estimates the survival function through the EM algorithm while excluding the phenotypic information of the proband (the individual that causes the family to be selected) to correct for ascertainment bias.

This method can handle ungenotyped individuals and allows the inclusion of covariates (such as sex, mutation type, etc.) in survival analysis, thus providing a more accurate estimate of the survival function. The paper also verifies the effectiveness of this method through simulated data and actual data sets (including family samples related to hereditary transthyretin - related amyloidosis).