-
Notifications
You must be signed in to change notification settings - Fork 0
/
Scenario_2_Analysis.R
243 lines (189 loc) · 15.1 KB
/
Scenario_2_Analysis.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
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
#Data analysis for Scenario 2
#To see the code for the final figures and PCA, see the Final_Figures file
setwd()
#Loading packages
library('ctv')
library('phytools')
library(ape)
library(phangorn)
library(seqinr)
library(adegenet)
library(tidyverse)
library(TreeSim)
library(apTreeshape)
library(phytools)
library('treebalance')
library('phyloTop')
devtools::install_github('Leonardini/treeCentrality')
library('treeCentrality')
library(devtools)
library(plyr)
#install_github("vqv/ggbiplot")
library(ggbiplot)
library(ggplot2)
library(gridExtra)
library(ggpmisc)
library(tibble)
library(magrittr)
library(stringr)
library('gginnards')
#Loading empirical trees
load(file="finaltreeparams.RData")
#Excluding trees with pendants that extend all the way (i.e., those with potential outgroups)--------------------------------
#Use a which statement to exclude trees with potential outgroups
nooutgroup <- filter(finaltreeparams, longestpendant != age, EmpOrSim == "emp")
treenums <- nooutgroup$treenumber
nooutgrouptrees <- finaltreeparams %>% filter(treenumber %in% treenums)
nooutgrouptrees
#Wilcoxon Rank Sum Test to compare empirical and simulated trees; no need to simulate new trees before analysis
library('dplyr')
empirical <- filter(nooutgrouptrees, EmpOrSim == "emp")
simulated <- filter(nooutgrouptrees, EmpOrSim == "sim")
wil_TotalI <- wilcox.test(empirical$TotalI, simulated$TotalI, paired = FALSE, alternative = "two.sided")
wil_MeanI <- wilcox.test(empirical$MeanI, simulated$MeanI, paired = FALSE, alternative = "two.sided")
wil_MedianI <- wilcox.test(empirical$MedianI, simulated$MedianI, paired = FALSE, alternative = "two.sided")
wil_NormColless <- wilcox.test(empirical$NormColless, simulated$NormColless, paired = FALSE, alternative = "two.sided")
wil_CollessLikeExpMDM <- wilcox.test(empirical$CollessLikeExpMDM, simulated$CollessLikeExpMDM, paired = FALSE, alternative = "two.sided")
wil_I2 <- wilcox.test(empirical$I2, simulated$I2, paired = FALSE, alternative = "two.sided")
wil_NormSackin <- wilcox.test(empirical$NormSackin, simulated$NormSackin, paired = FALSE, alternative = "two.sided")
wil_TotalCophenetic <- wilcox.test(empirical$TotalCophenetic, simulated$TotalCophenetic, paired = FALSE, alternative = "two.sided")
wil_B1 <- wilcox.test(empirical$B1, simulated$B1, paired = FALSE, alternative = "two.sided")
wil_B2 <- wilcox.test(empirical$B2, simulated$B2, paired = FALSE, alternative = "two.sided")
wil_AvgLeafDepth <- wilcox.test(empirical$AvgLeafDepth, simulated$AvgLeafDepth, paired = FALSE, alternative = "two.sided")
wil_leafdepthvariance <- wilcox.test(empirical$leafdepthvariance, simulated$leafdepthvariance, paired = FALSE, alternative = "two.sided")
wil_CPrank <- wilcox.test(empirical$CPrank, simulated$CPrank, paired = FALSE, alternative = "two.sided")
wil_normcherry <- wilcox.test(empirical$normcherry, simulated$normcherry, paired = FALSE, alternative = "two.sided")
wil_Jindex <- wilcox.test(empirical$Jindex, simulated$Jindex, paired = FALSE, alternative = "two.sided")
wil_SNI <- wilcox.test(empirical$SNI, simulated$SNI, paired = FALSE, alternative = "two.sided")
wil_Sshape <- wilcox.test(empirical$Sshape, simulated$Sshape, paired = FALSE, alternative = "two.sided")
wil_APP <- wilcox.test(empirical$APP, simulated$APP, paired = FALSE, alternative = "two.sided")
wil_normILnumber <- wilcox.test(empirical$normILnumber, simulated$normILnumber, paired = FALSE, alternative = "two.sided")
wil_normaverageladder <- wilcox.test(empirical$normaverageladder, simulated$normaverageladder, paired = FALSE, alternative = "two.sided")
wil_maxdepth <- wilcox.test(empirical$maxdepth, simulated$maxdepth, paired = FALSE, alternative = "two.sided")
wil_maxwidth <- wilcox.test(empirical$maxwidth, simulated$maxwidth, paired = FALSE, alternative = "two.sided")
wil_maxwidthtodepth <- wilcox.test((empirical$maxwidth/empirical$maxdepth), (simulated$maxwidth/simulated$maxdepth), paired = FALSE, alternative = "two.sided")
wil_maxdiffwidth <- wilcox.test(empirical$maxdiffwidth, simulated$maxdiffwidth, paired = FALSE, alternative = "two.sided")
wil_meandepth <- wilcox.test(empirical$meandepth, simulated$meandepth, paired = FALSE, alternative = "two.sided")
wil_maxheight <- wilcox.test(empirical$maxheight, simulated$maxheight, paired = FALSE, alternative = "two.sided")
wil_diameter <- wilcox.test(empirical$diameter, simulated$diameter, paired = FALSE, alternative = "two.sided")
wil_normpitchforks <- wilcox.test(empirical$normpitchforks, simulated$normpitchforks, paired = FALSE, alternative = "two.sided")
wil_rootedquartet <- wilcox.test(empirical$rootedquartet, simulated$rootedquartet, paired = FALSE, alternative = "two.sided")
wil_furnas <- wilcox.test(empirical$furnas, simulated$furnas, paired = FALSE, alternative = "two.sided")
wil_stairs <- wilcox.test(empirical$stairs, simulated$stairs, paired = FALSE, alternative = "two.sided")
wil_stairs2 <- wilcox.test(empirical$stairs2, simulated$stairs2, paired = FALSE, alternative = "two.sided")
wil_longestpendant <- wilcox.test(empirical$longestpendant, simulated$longestpendant, paired = FALSE, alternative = "two.sided")
wil_shortestpendant <- wilcox.test(empirical$shortestpendant, simulated$shortestpendant, paired = FALSE, alternative = "two.sided")
wil_stemminess <- wilcox.test(empirical$stemminess, simulated$stemminess, paired = FALSE, alternative = "two.sided")
wil_all_nooutgroup <- list(wil_TotalI, wil_MeanI, wil_MedianI, wil_NormColless, wil_CollessLikeExpMDM, wil_I2, wil_NormSackin,
wil_TotalCophenetic, wil_B1, wil_B2, wil_AvgLeafDepth, wil_leafdepthvariance, wil_CPrank, wil_normcherry, wil_Jindex, wil_SNI,
wil_Sshape, wil_APP, wil_normILnumber, wil_normaverageladder, wil_maxdepth,
wil_maxwidth, wil_maxdiffwidth, wil_meandepth, wil_maxheight,
wil_furnas, wil_normpitchforks, wil_rootedquartet, wil_stairs, wil_stairs2, wil_longestpendant,
wil_shortestpendant, wil_stemminess)
save(wil_all_nooutgroup, file="wilcoxon_values_nooutgroup.RData")
load(file="wilcoxon_values_nooutgroup.RData")
#Creating a data frame with all the Wilcoxon distance values and p-values
wilcoxon_distance_values_nooutgroup <- data.frame(
name=c("Total I","Mean I", "Median I", "Norm. Colless", "Colless-like index",
"I2", "Norm. Sackin", "Total coph. index", "B1", "B2", "Avg. leaf depth",
"Leaf depth variance", "Norm. cherries", "J index", "SNI", "S-shape statistic",
"APP", "Norm. IL number", "Norm. avg. ladder length", "Max. depth", "Max. width",
"Max. diff. in widths", "Mean depth", "Max. height",
"Norm. pitchforks", "Rooted Quartet", "Furnas rank",
"stairs", "stairs2", "Longest pendant edge", "Shortest pendant edge", "Stemminess"),
distance_value=c(wil_TotalI$statistic, wil_MeanI$statistic, wil_MedianI$statistic, wil_NormColless$statistic, wil_CollessLikeExpMDM$statistic,
wil_I2$statistic, wil_NormSackin$statistic,
wil_TotalCophenetic$statistic, wil_B1$statistic, wil_B2$statistic, wil_AvgLeafDepth$statistic, wil_leafdepthvariance$statistic,
wil_normcherry$statistic, wil_Jindex$statistic, wil_SNI$statistic,
wil_Sshape$statistic, wil_APP$statistic, wil_normILnumber$statistic, wil_normaverageladder$statistic, wil_maxdepth$statistic,
wil_maxwidth$statistic, wil_maxdiffwidth$statistic, wil_meandepth$statistic, wil_maxheight$statistic,
wil_normpitchforks$statistic, wil_rootedquartet$statistic, wil_furnas$statistic, wil_stairs$statistic, wil_stairs2$statistic,
wil_longestpendant$statistic,
wil_shortestpendant$statistic, wil_stemminess$statistic),
p_value=c(wil_TotalI$p.value, wil_MeanI$p.value, wil_MedianI$p.value, wil_NormColless$p.value, wil_CollessLikeExpMDM$p.value,
wil_I2$p.value, wil_NormSackin$p.value,
wil_TotalCophenetic$p.value, wil_B1$p.value, wil_B2$p.value, wil_AvgLeafDepth$p.value, wil_leafdepthvariance$p.value,
wil_normcherry$p.value, wil_Jindex$p.value, wil_SNI$p.value,
wil_Sshape$p.value, wil_APP$p.value, wil_normILnumber$p.value, wil_normaverageladder$p.value, wil_maxdepth$p.value,
wil_maxwidth$p.value, wil_maxdiffwidth$p.value, wil_meandepth$p.value, wil_maxheight$p.value,
wil_normpitchforks$p.value, wil_rootedquartet$p.value, wil_furnas$p.value, wil_stairs$p.value, wil_stairs2$p.value, wil_longestpendant$p.value,
wil_shortestpendant$p.value, wil_stemminess$p.value))
save(wilcoxon_distance_values_nooutgroup, file="wilcoxon_distance_values_nooutgroup.RData")
load(file="wilcoxon_distance_values_nooutgroup.RData")
#Calculating the z-score for each index, where the mean and SD are taken from the simulated distribution but the z-score is calculated for the empirical tree
zscores <- list("age", "gamma", "TotalI","MeanI", "MedianI", "NormColless", "CollessLikeExpMDM",
"I2", "Sackin", "NormSackin", "TotalCophenetic", "B1", "B2", "AvgLeafDepth",
"leafdepthvariance", "CPrank", "normcherry", "Jindex", "SNI", "Sshape",
"APP", "normILnumber", "normaverageladder", "maxdepth", "maxwidth",
"maxdiffwidth", "meandepth", "maxheight",
"diameter", "normpitchforks", "rootedquartet", "furnas",
"stairs", "stairs2", "longestpendant", "shortestpendant", "longestpendanttt", "stemminess")
#For loop to get the z-score values
i <- 1
for (i in i:1189) {
print(i)
{
tree2 <- NULL
tree2 <- filter(finaltreeparams, EmpOrSim == "emp", treenumber == i)
tree <- NULL
tree <- filter(finaltreeparams, EmpOrSim == "sim", treenumber == i)
zscores$age <- c(zscores$age,tree2$age)
zscores$longestpendanttt <- c(zscores$longestpendanttt,tree2$longestpendant)
zscores$gamma <- c(zscores$gamma,(tree2$gamma-(mean(tree$gamma)))/sd(tree$gamma))
zscores$TotalI <- c(zscores$TotalI,(tree2$TotalI-(mean(tree$TotalI)))/sd(tree$TotalI))
zscores$MeanI <- c(zscores$MeanI,(tree2$MeanI-(mean(tree$MeanI)))/sd(tree$MeanI))
zscores$MedianI <- c(zscores$MedianI,(tree2$MedianI-(mean(tree$MedianI)))/sd(tree$MedianI))
zscores$NormColless <- c(zscores$NormColless,(tree2$NormColless-(mean(tree$NormColless)))/sd(tree$NormColless))
zscores$CollessLikeExpMDM <- c(zscores$CollessLikeExpMDM,(tree2$CollessLikeExpMDM-(mean(tree$CollessLikeExpMDM)))/sd(tree$CollessLikeExpMDM))
zscores$I2 <- c(zscores$I2,(tree2$I2-(mean(tree$I2)))/sd(tree$I2))
zscores$Sackin <- c(zscores$Sackin,(tree2$Sackin-(mean(tree$Sackin)))/sd(tree$Sackin))
zscores$NormSackin <- c(zscores$NormSackin,(tree2$NormSackin-(mean(tree$NormSackin)))/sd(tree$NormSackin))
zscores$TotalCophenetic <- c(zscores$TotalCophenetic,(tree2$TotalCophenetic-(mean(tree$TotalCophenetic)))/sd(tree$TotalCophenetic))
zscores$B1 <- c(zscores$B1,(tree2$B1-(mean(tree$B1)))/sd(tree$B1))
zscores$B2 <- c(zscores$B2,(tree2$B2-(mean(tree$B2)))/sd(tree$B2))
zscores$AvgLeafDepth <- c(zscores$AvgLeafDepth,(tree2$AvgLeafDepth-(mean(tree$AvgLeafDepth)))/sd(tree$AvgLeafDepth))
zscores$leafdepthvariance <- c(zscores$leafdepthvariance,(tree2$leafdepthvariance-(mean(tree$leafdepthvariance)))/sd(tree$leafdepthvariance))
zscores$CPrank <- c(zscores$CPrank,(tree2$CPrank-(mean(tree$CPrank)))/sd(tree$CPrank))
zscores$normcherry <- c(zscores$normcherry,(tree2$normcherry-(mean(tree$normcherry)))/sd(tree$normcherry))
zscores$Jindex <- c(zscores$Jindex,(tree2$Jindex-(mean(tree$Jindex)))/sd(tree$Jindex))
zscores$SNI <- c(zscores$SNI,(tree2$SNI-(mean(tree$SNI)))/sd(tree$SNI))
zscores$Sshape <- c(zscores$Sshape,(tree2$Sshape-(mean(tree$Sshape)))/sd(tree$Sshape))
zscores$APP <- c(zscores$APP,(tree2$APP-(mean(tree$APP)))/sd(tree$APP))
zscores$normILnumber <- c(zscores$normILnumber,(tree2$normILnumber-(mean(tree$normILnumber)))/sd(tree$normILnumber))
zscores$normaverageladder <- c(zscores$normaverageladder,(tree2$normaverageladder-(mean(tree$normaverageladder)))/sd(tree$normaverageladder))
zscores$maxdepth <- c(zscores$maxdepth,(tree2$maxdepth-(mean(tree$maxdepth)))/sd(tree$maxdepth))
zscores$maxwidth <- c(zscores$maxwidth,(tree2$maxwidth-(mean(tree$maxwidth)))/sd(tree$maxwidth))
zscores$maxdiffwidth <- c(zscores$maxdiffwidth,(tree2$maxdiffwidth-(mean(tree$maxdiffwidth)))/sd(tree$maxdiffwidth))
zscores$meandepth <- c(zscores$meandepth,(tree2$meandepth-(mean(tree$meandepth)))/sd(tree$meandepth))
zscores$maxheight <- c(zscores$maxheight,(tree2$maxheight-(mean(tree$maxheight)))/sd(tree$maxheight))
zscores$averagepath <- c(zscores$averagepath,(tree2$averagepath-(mean(tree$averagepath)))/sd(tree$averagepath))
zscores$normpitchforks <- c(zscores$normpitchforks,(tree2$normpitchforks-(mean(tree$normpitchforks)))/sd(tree$normpitchforks))
zscores$rootedquartet <- c(zscores$rootedquartet,(tree2$rootedquartet-(mean(tree$rootedquartet)))/sd(tree$rootedquartet))
zscores$furnas <- c(zscores$furnas,(tree2$furnas-(mean(tree$furnas)))/sd(tree$furnas))
zscores$stairs <- c(zscores$stairs,(tree2$stairs-(mean(tree$stairs)))/sd(tree$stairs))
zscores$stairs2 <- c(zscores$stairs2,(tree2$stairs2-(mean(tree$stairs2)))/sd(tree$stairs2))
zscores$shortestpendant <- c(zscores$shortestpendant,(tree2$shortestpendant-(mean(tree$shortestpendant)))/sd(tree$shortestpendant))
zscores$stemminess <- c(zscores$stemminess,(tree2$stemminess-(mean(tree$stemminess)))/sd(tree$stemminess))
}
}
#Putting all the z-scores in a data frame
zscoresdf <- data.frame(zscores$age, zscores$longestpendanttt, zscores$gamma, zscores$TotalI, zscores$MeanI, zscores$MedianI, zscores$NormColless,
zscores$CollessLikeExpMDM, zscores$I2, zscores$Sackin, zscores$NormSackin, zscores$TotalCophenetic, zscores$B1,
zscores$B2, zscores$AvgLeafDepth,
zscores$leafdepthvariance, zscores$CPrank, zscores$normcherry, zscores$Jindex, zscores$SNI, zscores$Sshape,
zscores$APP, zscores$normILnumber,
zscores$normaverageladder, zscores$maxdepth, zscores$maxwidth,
zscores$maxdiffwidth, zscores$meandepth, zscores$maxheight,
zscores$normpitchforks, zscores$rootedquartet, zscores$furnas,
zscores$stairs, zscores$stairs2, zscores$shortestpendant, zscores$stemminess)
colnames(zscoresdf) <- c("age", "longestpendanttt", "gamma", "TotalI","MeanI", "MedianI", "NormColless", "CollessLikeExpMDM",
"I2", "Sackin", "NormSackin", "TotalCophenetic", "B1", "B2", "AvgLeafDepth",
"leafdepthvariance", "CPrank", "normcherry", "Jindex", "SNI", "Sshape",
"APP", "normILnumber", "normaverageladder", "maxdepth", "maxwidth",
"maxdiffwidth", "meandepth", "maxheight",
"normpitchforks", "rootedquartet", "furnas",
"stairs", "stairs2", "shortestpendant", "stemminess")
zscoresdf_nooutgroupdf <- filter(zscoresdf, longestpendanttt != age)
save(zscoresdf_nooutgroupdf, file="zscoresdf_nooutgroup.RData")
load(file="zscoresdf_nooutgroup.RData")
#All the final figures used in the paper are coded in the file: Final_Figures