Find differential exon usage, etc.

edger_group_confects(fit, group_id, group_effect, fdr = 0.05, step = 0.01)
limma_group_confects(object, group_id, group_effect, fdr = 0.05,
step = 0.01, trend = FALSE)

## Arguments

fit |
For edgeR, an edgeR DGEGLM object. |

group_id |
A factor of length `nrow(fit)` , assigning items to groups (eg genes). |

group_effect |
A group effect object created by one of the A `group_effect_...` functions. |

fdr |
False Discovery Rate to maintain. |

step |
Step size when calculating confident effect sizes. |

object |
For limma, an expression matrix or EList object, or anything limma's lmFit will accept. |

trend |
For limma, should `eBayes(trend=TRUE)` be used? |

## Value

See `nest_confects`

for details of how to interpret the result.

## Details

If the order of the members of a group is important, ensure the order of rows is correct in the original matrix.

Groups with less than two members will be ignored.

To construct a group design matrix, the design from `fit`

will be repeated in a block diagonal matrix.