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

- New function,
`subset_variants`

, which retains only variants with data bearing upon pathogenicity. - Return posterior mean of
`omega`

even when not explicitly sampled in`summary.BeviMed_m`

.

`bevimed_polytomous`

function added which enables application of BeviMed across multiple association models.`BeviMed`

objects now more general, representing results of inference with respect to the baseline model`gamma = 0`

and an arbitrary number of alternative association models - typically, one for each mode of inheritance. The`$moi`

slot has been replaced with`$models`

.`prob_pathogenic`

now returns a list when broken down by mode of inheritance/model.

- Make BeviMed work smoothly when number of individuals or number of variants is 0.
- Retain names of variants from columns of original allele count matrix.
- Improvements to guide, with more detail on model selection.

- Fixed bug in calculation of expected number of explaining variants by only including those with pathogenic configurations.

- Previous
`bevimed`

function now replaced by`bevimed_m`

, with the`_m`

indicating that it conditions on mode of inheritance. `bevimed`

now integrates over indicator of association (gamma) and mode of inheritance (m), allowing user to specify priors on probability of association and probability of dominance.- The
`BeviMed`

class object has been replaced by`BeviMed_m`

, and a new`BeviMed`

class has been introduced for inference with respect to all models: gamma 0 and gamma 1 under each mode of inheritance. - A new vignette with more detail called
`BeviMed Guide`

which relates the package to the paper. - Names used for summary statistics in summary objects have changed, see function help pages for details on current names.
`print`

ing a`BeviMed`

object now shows conditional probabilities of pathogenicity for each mode of inheritance, and expected explained cases and expected explaining variants shown too.- Bug fixed in adaptive tuning for omega and phi proposals.

- Re-naming of parameters in
`bevimed`

function to match the names of variables in the paper (under submission). - The allele count matrix
`G`

should now be supplied as a matrix with rows corresponding to individuals, not variants. `expected_explained`

and`explaining_variants`

functions 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.