-
Notifications
You must be signed in to change notification settings - Fork 0
/
clade_manip_LowExtinction.R
190 lines (158 loc) · 8.27 KB
/
clade_manip_LowExtinction.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
library(geiger)
library(caper)
library(pez)
library(phytools)
library(moments)
library(apTreeshape)
library(parallel)
find.clade <- function(tree, tips)
{
clade.mat <- clade.matrix(tree)$clade.matrix
sums <- rowSums(clade.mat)
ancestors <- apply(clade.mat, 1, function(x) all(x[tips]==1))
sums[!ancestors] <- length(tree$tip.label)
return(unname(which.min(sums)))
}
bind.replace <- function(backbone, donor, replacing.tip.label)
{
size <- max(branching.times(donor))
bind.point <- which(backbone$tip.label == replacing.tip.label)
tree <- bind.tree(backbone, donor, where=bind.point)
clade <- find.clade(tree, which(tree$tip.label %in% donor$tip.label))
which.edge <- which(tree$edge[,2] == clade)
if(tree$edge.length[which.edge] < size)
stop("Donor phylogeny too large for backbone phylogeny")
tree$edge.length[which.edge] <- tree$edge.length[which.edge] - size
return(tree)
}
get.ed <- function(spp, size)
{
#Running the function in order to get the ED value correlations for the focal clades.
size.met <- FALSE
while (!size.met)
{
tree <- sim.bdtree(b = 1, d = 0.5, n = spp, stop = "taxa", extinct=FALSE)
tree <- drop.extinct(tree)
clade.mat <- clade.matrix(tree)$clade.matrix
mat.sums <- rowSums(clade.mat)
size.met <- any(mat.sums == size)
}
selected.clades <- which(mat.sums == size)
if (length(selected.clades) == 1){
clade.number <- selected.clades
}else{
clade.number<-sample(selected.clades, 1)
}
#Getting the spp of the focal clade so that the ED values of these spp can be selected out of
#original.ed and imputed.ed.
random.clade <- extract.clade(tree, clade.number)
focal.clade.spp <- unique(na.omit(as.numeric(unlist(strsplit(unlist(random.clade$tip.label), "[^0-9]+")))))
#calculate the ED of the original tree
original.ed <- ed.calc(tree)$spp
original.ed <- setNames(original.ed[,2], original.ed[,1])
# Rank the spp in the original tree upon their ED values
original.ranking <- rank(original.ed)
# Get the ED values of all spp excluding the random clade which has been selected for removal.
excluding.clade.original <- original.ed[!names(original.ed) %in% random.clade$tip.label]
# Get the ED values of all spp in the random clade which has been selected for removal.
focal.clade.original.ed <- original.ed[names(original.ed) %in% random.clade$tip.label]
# Get the true ranking of spp in the clade to be removed.
original.clade.ranking <- original.ranking[names(original.ranking) %in% random.clade$tip.label]
# Get the true ranking of spp outside of the focal clade.
original.excluding.ranking <- original.ranking[!names(original.ranking) %in% random.clade$tip.label]
# Calculate the average ED of the clade to be removed.
original.clade.averageED <- mean(focal.clade.original.ed)
# Replace the ED scores of the spp in the clade to be removed with the average ED score of that clade.
meanClade.original.ed <- original.ed
meanClade.original.ed[names(meanClade.original.ed) %in% random.clade$tip.label] <- original.clade.averageED
# Rank all spp again while the ED scores for spp in focal clade are the avg of that focal clade.
mean.original.ranking <- rank(meanClade.original.ed)
# Get ranking of spp in focal clade which have avg ED score
original.meanClade.ranking <- mean.original.ranking[names(mean.original.ranking) %in% random.clade$tip.label]
node.name <- which(names(branching.times(tree)) == clade.number)
# Compare the original clade rankings to the rankings if the avg ED were
# applied to each of the spp in the focal clade.
average.ed.comparison <- sum(abs(original.clade.ranking - original.meanClade.ranking))/size
#Calculating for original.tree
original.lambda <- yule(tree)$lambda
node.age <- branching.times(tree)[node.name]
nasty.bl <- dist.nodes(tree)[clade.number,][clade.number-1]
gammatree <- ltt(tree, plot = FALSE, gamma = FALSE)
original.gamma <- gammatest(gammatree)$gamma
colless.tree <- as.treeshape(tree, model = "yule")
original.colless <- colless(colless.tree)
original.kurtosis <- kurtosis(original.ed)
original.skew <- skewness(original.ed)
original.sd <- sd(original.ed)
original.branches <- sum(tree$edge.length)
dropped.tree <- drop.tip(tree, sample(random.clade$tip.label, length(random.clade$tip.label)-1))
r <- NULL
while(is.null(r)){
try({
donor.clade <- sim.bdtree(b = 1, d = 0.5, n = size, stop = "taxa", extinct=FALSE)
donor.clade <- drop.extinct(donor.clade)
donor.clade$tip.label <- random.clade$tip.label
for (i in dropped.tree$tip.label){
if(any(random.clade$tip.label==i)){
tip <- i
}
}
r <- bind.replace(dropped.tree, donor.clade, tip)
dropped.tree <- r
})
}
#calculating for imputed.tree
imputed.colless.tree <- as.treeshape(dropped.tree, model = "yule")
imputed.colless <- colless(imputed.colless.tree)
imputed.lambda <- yule(dropped.tree)$lambda
imputed.gammatree<-ltt(dropped.tree, plot = FALSE, gamma = FALSE)
imputed.gamma <- gammatest(imputed.gammatree)$gamma
imputed.ed <- ed.calc(dropped.tree)$spp
imputed.ed <- setNames(imputed.ed[,2], imputed.ed[,1])
imputed.ranking <- rank(imputed.ed)
imputed.kurtosis <- kurtosis(imputed.ed)
imputed.skew <- skewness(imputed.ed)
imputed.sd <- sd(imputed.ed)
imputed.branches <- sum(dropped.tree$edge.length)
#calculating for imputed clade (donor.clade)
imputed.clade.lambda <- yule(donor.clade)$lambda
original.clade.lambda <- yule(random.clade)$lambda
imputed.clade.branches <- sum(donor.clade$edge.length)
original.clade.branches <- sum(random.clade$edge.length)
# Get the ED values of the tree excluding the spp on the imputed clade.
excluding.clade.imputed <- imputed.ed[!names(imputed.ed) %in% random.clade$tip.label]
# Get the rank of the spp in the tree excluding spp in the imputed clade.
excluding.clade.imputed.rank <- rank(excluding.clade.imputed)
#Compare the difference in ranks of spp outside of the focal clade.
remainingspp.ranking.error <- sum(abs(original.excluding.ranking-excluding.clade.imputed.rank))/(spp-size)
ranking.error <- sum(abs(original.ranking[random.clade$tip.label]-imputed.ranking[random.clade$tip.label]))/size
error.params <- c(50,100,200,0.05*spp,0.1*spp,0.2*spp)
error.rate <- NA
for (param in error.params){
error.rate[which(error.params == param)] <- sum(!(imputed.ranking[1:param] %in% original.ranking[1:param]))/param
}
full.ed.corr <- cor(excluding.clade.original, excluding.clade.imputed)
focal.ed.corr <- cor(original.ed[random.clade$tip.label], imputed.ed[random.clade$tip.label])
return(c(full.ed.corr, focal.ed.corr, original.gamma, imputed.gamma, original.lambda, imputed.lambda,
original.colless, imputed.colless, original.kurtosis, imputed.kurtosis, original.skew, imputed.skew,
original.sd, imputed.sd, imputed.clade.lambda, original.clade.lambda, original.branches, imputed.branches,
original.clade.branches, imputed.clade.branches, error.rate, ranking.error, original.clade.averageED,
average.ed.comparison, remainingspp.ranking.error))
}
wrapper <- function(n.spp=floor(2^seq(8,10,0.2)), clade.size=seq(6, 32, 2), reps = 5)
{
data <- expand.grid(n.spp = n.spp, clade.size = clade.size, reps=1:reps)
sim.wrap <- function(x)
return(get.ed(data$n.spp[x], data$clade.size[x]))
output <- do.call(rbind, mcMap(sim.wrap, seq_len(nrow(data)), mc.cores=12))
data <- cbind(data, output)
names(data)[-1:-3] <- c("full.ed", "focal.ed", "original.gamma", "imputed.gamma", "original.lambda", "imputed.lambda",
"original.colless", "imputed.colless", "original.kurtosis", "imputed.kurtosis", "original.skew",
"imputed.skew", "original.sd", "imputed.sd", "imputed.clade.lambda","original.clade.lambda",
"original.branches", "imputed.branches", "original.clade.branches", "imputed.clade.branches",
"error.rate.50", "error.rate.100" ,"error.rate.200", "error.rate.5pct", "error.rate.10pct",
"error.rate.20pct", "ranking.error", "Average.Clade.ED", "Average.Clade.ED.RankingError",
"RankingError.Remaining.Spp")
write.table(data, "correlations_LowExtinction.txt")
}
wrapper()