A fast integrative genetic association test for rare diseases based on a model for disease status given allele counts at rare variant sites. Probability of association, mode of inheritance and probability of pathogenicity for individual variants are all inferred in a Bayesian framework - 'A Fast Association Test for Identifying Pathogenic Variants Involved in Rare Diseases', Greene et al 2017
subset_variants, which retains only variants with data bearing upon pathogenicity.
omegaeven when not explicitly sampled in
bevimed_polytomousfunction added which enables application of BeviMed across multiple association models.
BeviMedobjects now more general, representing results of inference with respect to the baseline model
gamma = 0and an arbitrary number of alternative association models - typically, one for each mode of inheritance. The
$moislot has been replaced with
prob_pathogenicnow returns a list when broken down by mode of inheritance/model.
bevimedfunction now replaced by
bevimed_m, with the
_mindicating that it conditions on mode of inheritance.
bevimednow integrates over indicator of association (gamma) and mode of inheritance (m), allowing user to specify priors on probability of association and probability of dominance.
BeviMedclass object has been replaced by
BeviMed_m, and a new
BeviMedclass has been introduced for inference with respect to all models: gamma 0 and gamma 1 under each mode of inheritance.
BeviMed Guidewhich relates the package to the paper.
BeviMedobject now shows conditional probabilities of pathogenicity for each mode of inheritance, and expected explained cases and expected explaining variants shown too.
bevimedfunction to match the names of variables in the paper (under submission).
Gshould now be supplied as a matrix with rows corresponding to individuals, not variants.
explaining_variantsfunctions have been added, respectively computing the expected number of cases with their disease explained by the given variants, and expected number of pathogenic variants present amongst cases.