从系统发育树的集合中获得平均系统发育树分支长度

如果我理解得很好，您想将每个唯一的所有树分类到不同的组中（例如，在您的示例中，第一组包含一棵树等......）然后测量每个组的一些统计数据？

您可以通过首先将拓扑分组到列表中来做到这一点：

set.seed(5)
## Generating 20 4 tip trees (hopefully they will be identical topologies!)
tree_list <- rmtree(20, 4)
## How many unique topologies?
length(unique(tree_list))
## Sorting the trees by topologies
tree_list_tmp <- tree_list
sorted_tree_list <- list()
counter <- 0
while(length(tree_list_tmp) != 0) {
    counter <- counter+1
    ## Is the first tree equal to any of the trees in the list
    equal_to_tree_one <- unlist(lapply(tree_list_tmp, function(x, base) all.equal(x, base, use.edge.length = FALSE), base = tree_list_tmp[[1]]))
    ## Saving the identical trees
    sorted_tree_list[[counter]] <- tree_list_tmp[which(equal_to_tree_one)]
    ## Removing them from the list
    tree_list_tmp <- tree_list_tmp[-which(equal_to_tree_one)]
    ## Repeat while there are still some trees!
}
## The list of topologies should be equal to the number of unique trees
length(sorted_tree_list) == length(unique(tree_list))
## Giving them names for fancyness
names(sorted_tree_list) <- paste0("topology", 1:length(sorted_tree_list))

然后对于每个唯一拓扑组中的所有树，您可以通过创建函数来提取不同的汇总统计信息。例如，这里我将测量分支长度 sd、平均值和 90% 分位数。

## function for getting some stats
get.statistics <- function(unique_topology_group) {
    ## Extract the branch lengths of all the trees
    branch_lengths <- unlist(lapply(unique_topology_group, function(x) x$edge.length))

    ## Apply some statistics
    return(c(   n = length(unique_topology_group),
             mean = mean(branch_lengths),
               sd = sd(branch_lengths),
               quantile(branch_lengths, prob = c(0.05, 0.95))))
} 

## Getting all the stats
all_stats <- lapply(sorted_tree_list, get.statistics)
## and making it into a nice table
round(do.call(rbind, all_stats), digits = 3)
#            n  mean    sd    5%   95%
# topology1  3 0.559 0.315 0.113 0.962
# topology2  2 0.556 0.259 0.201 0.889
# topology3  4 0.525 0.378 0.033 0.989
# topology4  2 0.489 0.291 0.049 0.855
# topology5  2 0.549 0.291 0.062 0.882
# topology6  1 0.731 0.211 0.443 0.926
# topology7  3 0.432 0.224 0.091 0.789
# topology8  1 0.577 0.329 0.115 0.890
# topology9  1 0.473 0.351 0.108 0.833
# topology10 1 0.439 0.307 0.060 0.795

当然，您可以调整它以获得您自己想要的统计数据，甚至可以获取每组每棵树的统计数据（使用双重叠lapply(sorted_trees_list, lapply, get.statistics)或类似方法）。