###
#This file contains R code to reconstruct Figs 1-4, S1, S3-S4, S6 and S8-S11 from Krasovec et al. 2017.
#the nlme, ggplot2, ape and dplyr packages will need to be installed if not already present (e.g. using the following line without the # at the start:
#install.packages(c("nlme", "ggplot2", "ape", "dplyr"))

#Set the working directory by adjusting this line to indicate the folder containing the data (S1_Data.csv) and removing the # at the beginning:
#setwd("/Path/to/folder")
#the code may be run all together, or bit by bit from an R terminal

library(nlme)
library(ggplot2)
library(ape)
library(dplyr)

##create the Fig.1
# read in the data and re-order organisms phylogenetically
d1 <-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "1") %>%
  mutate(Microorganism = factor(organism, levels=c("Halobacterium salinarum","Haloferax volcanii","Sulfolobus acidocaldarius","Bacillus anthracis", "Staphylococcus aureus","Mycoplasma mycoides","Corynebacterium glutamicum","Mycobacterium smegmatis",
"Mycobacterium tuberculosis","Deinococcus radiodurans","Thermus thermophilus","Helicobacter pylori","Brucella abortus","Brucella melitensis","Pseudomonas aeruginosa","Pseudomonas fluorescens","Haemophilus influenzae","Yersinia pestis","Escherichia coli","Salmonella enterica Typhimurium","Salmonella enterica Enteritidis","Saccharomyces cerevisiae","Bacteriophage phiX174","Bacteriophage Qbeta","Measles virus","Vesicular stomatitis virus")))

#create appropriate organism names
l <- levels(d1$Microorganism)
l[match(c("Bacteriophage phiX174","Bacteriophage Qbeta","Measles virus", "Vesicular stomatitis virus"),l)] <- c(expression(paste("Bacteriophage ", phi, "X174")), expression(paste("Bacteriophage Q", beta)), expression(paste("Measles virus")), expression(paste("Vesicular stomatitis virus")))

#choose appropriate colours
Archaea <- c("#691F01","#B13E0F", "#FF6600")
Bacteria <- c("#691F01","#B13E0F", "#FF6600", "#FF9955","#FFFF00","#ffff99","#A0A000","#BED661","#65D74D","#3AA655","#105952","#CBFFFA","#4997D0","#00468C","#002147","#d49dff","purple", "#4B0082")
Eukaryote <- c("#FF0000")
Virus <- c("#691F01","#B13E0F", "#FF6600", "#FF9955")

#create the figure
p1 <- ggplot(aes(x = D,y = mutation_rate,  fill = Microorganism, shape = Microorganism), data=d1)

graphics.off()
pdf("Fig1.pdf",width=11, height=7)

p1 + 
  geom_point(alpha=0.7, size=5) +
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), breaks=c(0.1,10,1000,100000), labels=c("0.1", "10", "1000", "100000"))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), breaks = 10^seq(4,10, by=2), labels=c(expression(10^4),expression(10^6),expression(10^8),expression(10^10))) +
  scale_shape_manual(name = "",values = c(rep(21, 3), rep(24, 18), 22, rep(23,4)), labels=l) +
  scale_fill_manual(name = "", values = c(Archaea,Bacteria, Eukaryote, Virus), labels=l) +
  theme(legend.position = c(0.25,0.27), 
        legend.key.size = unit(0.55,"cm"),
        legend.title = element_text(face = "bold", size = 14),
        legend.title.align = 0.5, 
        legend.text = element_text(face = "italic", size = 10), 
        legend.key = element_rect(fill = 'white'), 
        legend.text.align = 0, 
        panel.background = element_rect(fill = "white"), 
        axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.title.x = element_text(size=14), 
        axis.title.y = element_text(size=14), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()


##calibration model (Model S-IV) used below to convert LUM measurements to D
dc <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  filter(organism %in% c("Escherichia coli", "Pseudomonas aeruginosa"), !is.na(LUM))

modCalib <- lme(log2(D)~I(log2(LUM)-21.816664392518216)*organism, data=dc, random=list(~I(log2(LUM)-21.816664392518216)|genotype, ~1|block, ~1|plate_ID),  weights=varComb(varIdent(form=~1|block), varPower(form=~fitted(.))))

## Fig. 2A; first create Model S-II
# read in the data and create necessary composite columns
d2a <-read.csv("S1_Data.csv") %>%
  subset(figure == "2a") %>%
  mutate(Cal = 2^predict(modCalib, newdata=., level=1), 
         CalCent=log2(Cal)-mean(log2(Cal)), 
         genotype_marker=factor(paste(genotype, marker, sep = '_')),
         organism=relevel(organism, ref="Escherichia coli"),
         genotype_marker=relevel(genotype_marker, ref="MG1655_rifampicin50"),
         sep_2a=factor(paste(included, marker, sep = '_')),
         organism=(factor(organism)))

#create the model
mod2a <- lme(log2(mutation_rate)~CalCent + genotype_marker + CalCent:organism, data=d2a, random=~1|researcher/block/plate_ID, weights=varPower(form = ~SD_m*C_cultures))

#use model to create values for the lines in the figure
d2a$pred <- predict(mod2a,level=0)

#create the figure
p2a <- ggplot(aes(x = Cal,y = mutation_rate,  colour = organism, shape = sep_2a), data=d2a)

pdf("Fig2a.pdf",width=4, height=4.4)

p2a + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), (breaks=c(0.3,2,20)))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(4e7, 1e8, 5e8))) +
  scale_shape_manual(values=c(2,19,17), guide=FALSE) +
  scale_color_manual(values=c("#000A44","#35A2AD"),guide=FALSE) +
  geom_line(aes(group=genotype_marker, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()

## Fig. 2B; first create Model S-III
# read in the data and create necessary composite columns
d2b <- read.csv("S1_Data.csv") %>%
  filter(figure %in% c("2b"), !is.na(CC)) %>%
  mutate(l2Ccent=log2(CC)-mean(log2(CC)), 
         genotype=factor(genotype), genotype=relevel(factor(genotype), ref="BY4742"),
         sep_2b=factor(paste(included, marker, sep = '_')),
         gen2=(factor(genotype)))

levels(d2b$gen2) <- c("BY4742","S288_&_Sigma1278b","S288_&_Sigma1278b")

#create model
mod2b <- lme(log2(mutation_rate)~l2Ccent + genotype + gen2:l2Ccent, data=d2b, random=~1|block/plate_ID, weights=varComb(varIdent(form=~1|block), varPower(form=~culture_volume)))

#use model to create values for the lines in the figure
d2b$pred <- predict(mod2b,level=0)

#create the figure
p2b <- ggplot(aes(x = CC, y = mutation_rate,  colour = genotype, shape = sep_2b, fill=genotype), data=d2b)

pdf("Fig2b.pdf", width=4, height=4.4)

p2b + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), (breaks=c(10, 100, 1000, 5000)))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(1e6, 1e7, 1e8))) +
  scale_shape_manual(values=c(5,0,23,22), guide=FALSE) +
  scale_color_manual(values=c("#660000","#FF0000","#FF9933"), guide=FALSE) +
  scale_fill_manual(values=c("#660000","#FF0000","#FF9933"), guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()

## Fig. 3; first create Model S-VII
# read in the data and create necessary composite columns
d3 <-read.csv("S1_Data.csv", header = TRUE) %>%
  subset(figure == "3") %>%
  mutate(Cal = 2^predict(modCalib, newdata=., level=1), 
         CalCent=log2(Cal)-mean(log2(Cal)),
         system=(factor(genotype)),
         system2=(factor(genotype)),
         genotype = factor(genotype),
         marker = factor(marker))
  levels(d3$system) <- c("dam","dinB","metI","MMR","MMR","MMR","nei") 
  levels(d3$system2) <- c("damBmetI","damBmetI","damBmetI","MMR","MMR","MMR","nei")
  levels(d3$genotype) <- c("mutL", "mutH", "mutS", "nei", "metI", "dinB","dam")
  levels(d3$marker) <- c("nalidixic_acid30", "rifampicin50")
  d3 <- arrange(d3, marker,genotype)

  #create the model
mod3 <- lme(log2(mutation_rate)~CalCent + system + CalCent:system2, data=d3, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~upper_bound)))

#use model to create values for the lines in the figure
d3$pred <- predict(mod3,level=0) 

#create the figure
p3 <- ggplot(aes(x = Cal,y = mutation_rate,  colour = genotype, shape = marker), data=d3)

pdf("Fig3.pdf",width=4.4, height=4.4)

p3 + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), (breaks=c(10,40, 100)))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(5e7, 1e8,5e8))) +
  scale_shape_manual(values=c(19,17), guide=FALSE) +
  scale_color_manual(values=c("#E69F00","#CC79A7","#0CEE0C","#0072B2","#000A44", "#4DCDFF","#075607"), guide=FALSE) +
  geom_line(aes(y = 2^pred), alpha=1, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14),
        plot.margin=unit(c(5.5, 15, 5.5, 5.5), "points")) 
graphics.off()

##Fig. 4A; first create Model S-VIII
# read in the data and create necessary columns
d4a <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4a") %>%
  mutate(Cal = 2^predict(modCalib, newdata=., level=1), 
         CalCent=log2(Cal)-mean(log2(Cal)))
d4a$genotype=relevel(d4a$genotype, ref="mutT1")

#create model
mod4a <- lme(log2(mutation_rate)~1 + genotype, data=d4a, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~upper_bound)))

#use model to create values for the lines in the figure
d4a$pred <- predict(mod4a,level=0)

#create the figure
p4a <- ggplot(aes(x = Cal,y = mutation_rate,  colour = genotype, shape = marker), data=d4a)

pdf("Fig4a.pdf",width=4, height=4.4)
p4a + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), limits=c(5,95), breaks=c(6,30,90))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(6e7, 3e8, 6e8))) +
  scale_color_manual(values=c("#246C7F","#4DCDFF"),guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4, shape=19) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()

##Fig. 4B; first create Model S-X
# read in the data and create necessary columns

d4b <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4b") %>%
  mutate(Cal = 2^predict(modCalib, newdata=., level=1), 
         CalCent=log2(Cal)-mean(log2(Cal)),
         genotype=relevel(factor(genotype), ref="mutM"))

#create model
mod4b <- lme(log2(mutation_rate)~CalCent+genotype, data=d4b, random=~1|block/plate_ID, weights=varPower(form=~SD_m|genotype))

#use model to create values for the lines in the figure
d4b$pred <- predict(mod4b,level=0)

#create the figure
p4b <- ggplot(aes(x = Cal,y = mutation_rate,  colour = genotype, shape = marker), data=d4b)
pdf("Fig4b.pdf",width=4, height=4.4)
p4b + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), limits=c(5,95), breaks=c(6,30,90))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(8e7, 1.5e8, 3e8))) +
  scale_color_manual(values=c("#251DCC","#6BD693"),guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4, shape=17) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()

#Fig. 4C; first create Model S-XII
# read in the data and create necessary columns
d4c <-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4c") %>%
  mutate(l2Dcent=log2(D)-mean(log2(D)), gen2=(factor(genotype)))
  
#create model
mod4c <- lme(log2(mutation_rate)~1, data=d4c, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~upper_bound)))

# use model to create values for the lines in the figure
d4c$pred <- predict(mod4c,level=0)

#create the figure
p4c <- ggplot(aes(x = D,y = mutation_rate,  colour = genotype, shape = included), data=d4c)
pdf("Fig4c.pdf",width=4, height=4.4)
p4c + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), breaks=c(500,3000, 12000))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(1e6,1e7,1e8))) +
  scale_shape_manual(values=c(0,15), guide=FALSE) +
  scale_color_manual(values=c("#FF671B","#663300"),guide=FALSE) +
  geom_line(aes(y = 2^pred), lwd=1, colour="#FF671B", alpha=0.8) +
  geom_line(aes(y = 2^pred), lwd=1, colour="#663300", alpha=0.8, lty=2) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()

#Fig. 4D; first create Model S-XIV
# read in the data and create necessary columns
d4d <-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4d") %>%
  mutate(l2Dcent=log2(D)-mean(log2(D)))

#create model
mod4d <- lme(log2(mutation_rate)~l2Dcent, data=d4d, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~lower_bound)))

# use model to create values for the lines in the figure
d4d$pred <- predict(mod4d,level=0)

#create the figure
p4d <- ggplot(aes(x = D,y = mutation_rate,  colour = genotype, shape = marker), data=d4d)
pdf("Fig4d.pdf",width=4, height=4.4)
p4d + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), breaks=c(2,10,30))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(1.5e8,2.5e8,4e8))) +
  scale_color_manual(values=c("#F2625A"),guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^d4d$pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4, shape=18) +
  theme_classic(base_size = 14) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()

#### S1 Fig; 
# read in the data
dS1 <- read.tree(text = "(((((((((((Escherichia_coli:0.02123,(Salmonella_enterica_Typhimurium:0.02012,Salmonella_enterica_Enteritidis:0.00491)Node11:0.00492)Node10:0.08652,Yersinia_pestis:0.05702)Node9:0.00494,Haemophilus_influenzae:0.17514)Node8:0.19559,(Pseudomonas_fluorescens:0.06599,Pseudomonas_aeruginosa:0.05154)Node12:0.11363)Node7:0.09275,(Brucella_melitensis:0,Brucella_abortus:0)Node13:0.23462)Node6:0.02723,Helicobacter_pylori:0.34978)Node5:0.02114,(Deinococcus_radiodurans:0.23191,Thermus_thermophilus:0.1286)Node14:0.08484)Node4:0.01698,((Mycobacterium_tuberculosis:0.07389,Mycobacterium_smegmatis:0.0134)Node16:0.05793,Corynebacterium_glutamicum:0.0972)Node15:0.23763)Node3:0.00799,((Mycoplasma_mycoides:0.28176,Staphylococcus_aureus:0.08621)Node18:0.03732,Bacillus_anthracis:0.05643)Node17:0.16158)Node2:0.19854,(((Haloferax_volcanii:0.11902,Halobacterium_salinarum:0.08265)Node20:0.28483,Sulfolobus_acidocaldarius:0.21144)Node19:0.04196646,Saccharomyces_cerevisiae:0.5992681)NA:0.10979354)Node1:0.8312,(Mononegavirales:0.0069621,Bacteriophage_phi:0.0069621)NA:0,Bacteriophage_Q_beta:0.0069621)NA:0.1688;")
dS1$tip.label = gsub("_", " ", dS1$tip.label)
dS1$tip.label[c(24,25)] <- c(expression(paste("Bacteriophage ", phi, "X174")), expression(paste("Bacteriophage Q", beta)))

pdf("S1_Fig.pdf",width=7, height=7)
plot(dS1)
graphics.off()

#### S3 Fig; first create Model S-IV
# read in the data and create necessary composite columns
dS3 <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  filter(organism %in% c("Escherichia coli", "Pseudomonas aeruginosa"), !is.na(LUM))

#create the model
modCalib <- lme(log2(D)~I(log2(LUM)-21.816664392518216)*organism, data=dS3, random=list(~I(log2(LUM)-21.816664392518216)|genotype, ~1|block, ~1|plate_ID),  weights=varComb(varIdent(form=~1|block), varPower(form=~fitted(.))))

#use model to create values for the lines in the figure
dS3$pred <- predict(modCalib,level=1)

#create the figure
pS3 <- ggplot(aes(x = LUM,y = D,  colour = genotype, shape = strain), data=dS3)

pdf("S3_Fig.pdf",width=11, height=7)

pS3 + 
  scale_y_log10(expression(paste("Population density (CFU)")), (breaks=c(3e7,1e8,3e8,6e8)))  + 
  scale_x_log10(expression(paste("Luminescence (AU)")), (breaks=c(1.12e6,3.2e6,1.01e7))) +
  scale_color_manual(values=c("#691F01","#B13E0F", "#FF9955","#FFFF00","#ffff99","#A0A000","#BED661","#65D74D",
                              "#105952","#4997D0","#00468C","#002147","#d49dff","purple", "#4B0082")) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()

#### S4A Fig; first create Model S-V
# read in the data and create necessary composite columns
dS4a <-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "2a") %>%
  mutate(l2Dcent=log2(D)-mean(log2(D)), genotype_marker=factor(paste(genotype, marker, sep = '_')),
         genotype_marker=relevel(genotype_marker, ref="PAO1_rifampicin50"),
         sep_2a=factor(paste(included, marker, sep = '_')),
         gen2=(factor(genotype_marker)))
levels(dS4a$gen2) <- c("PAO1","MG1655","MG1655")

#create the model
modS4a <- lme(log2(mutation_rate)~l2Dcent + genotype_marker + l2Dcent:gen2, data=dS4a, random=~1|researcher/block/plate_ID, weights=varComb(varPower(form=~m), varPower(form=~C_cultures)))

#use model to create values for the lines in the figure
dS4a$pred <- predict(modS4a,level=0)

#create the figure
pS4a <- ggplot(aes(x = D,y = mutation_rate,  colour = organism, shape = sep_2a), data=dS4a)

pdf("S4A_Fig.pdf",width=4, height=4.4)

pS4a + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), (breaks=c(0.3,2,20)))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(4e7, 2e8, 6e8))) +
  scale_shape_manual(values=c(2,19,17), guide=FALSE) +
  scale_color_manual(values=c("#000A44","#35A2AD"),guide=FALSE) +
  geom_line(aes(group=genotype_marker, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()

#### S4B Fig; first create Model S-VI
# read in the data and create necessary composite columns
dS4b <- read.csv("S1_Data.csv") %>%
  filter(figure %in% c("2b"), !is.na(D)) %>%
  mutate(l2Dcent=log2(D)-mean(log2(D)), 
         genotype=factor(genotype), genotype=relevel(factor(genotype), ref="BY4742"),
         sep_2b=factor(paste(included, marker, sep = '_')),
         gen2=(factor(genotype)))
levels(dS4b$gen2) <- c("BY4742","S288_&_Sigma1278b","S288_&_Sigma1278b")


#create the model
modS4b <- lme(log2(mutation_rate)~l2Dcent + genotype + gen2:l2Dcent, data=dS4b, random=~1|block/plate_ID, weights=varPower(form=~generation_time*culture_volume))

#use model to create values for the lines in the figure
dS4b$pred <- predict(modS4b,level=0)

#create the figure
pS4b2 <- ggplot(aes(x = D, y = mutation_rate,  colour = genotype, shape = sep_2b, fill=genotype), data=dS4b)

pdf("S4B_Fig.pdf",width=4, height=4.4)

pS4b2 +  
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), (breaks=c(10, 100, 1000, 5000)))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(1e6, 1e7, 1e8))) +
  scale_shape_manual(values=c(5,0,23,22), guide=FALSE) +
  scale_color_manual(values=c("#660000","#FF0000","#FF9933"), guide=FALSE) +
  scale_fill_manual(values=c("#660000","#FF0000","#FF9933"), guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x = element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14) ) 
graphics.off()



#### S6 Fig;
# read in the data and create necessary composite columns
dS6 <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  filter(figure %in% c("S6"))

#create the model
modS6 <- gls(W~glucose_mg_per_L*genotype, data=dS6)

#use model to create values for the lines in the figure
dS6$pred <- predict(modS6,level=0)

#create the figure
pS6 <- ggplot(aes(x = factor(glucose_mg_per_L) ,y = W,  colour = environment, shape=genotype), data=dS6)

pdf("S6_Fig.pdf",width=4, height=4.4)

pS6 + 
  scale_y_continuous(expression(paste("fitness")), (breaks=c(0.85,0.9,0.95,1)))  + 
  scale_x_discrete(expression(paste("glucose (mg per L)"))) +
  scale_color_manual(values=c("#d49dff","#4B0082")) +
  scale_shape_manual(values=c(17,19)) +
  geom_boxplot(outlier.shape = NA) +
  facet_grid(~genotype) +
  geom_point(alpha=0.6, size=2) +
  theme(legend.position="none", panel.grid.major = element_blank(), panel.grid.minor = element_blank(), 
        panel.background = element_rect(fill = "white"), panel.border = element_blank(),axis.line.x = element_line(color = 'black', size=1), 
        axis.line.y = element_line(color = 'black', size=1), axis.title.x = element_text(face="bold", size=14), 
        axis.title.y = element_text(face="bold", size=14), axis.ticks=element_line(color = 'black', size=2), 
        axis.text.x=element_text(color = 'black', size=14), 
        axis.text.y=element_text(color = 'black', size=14) ) 
graphics.off() 

#### S8A Fig; first create Model S-IX
# read in the data and create necessary composite columns
dS8a <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4a") %>%
  mutate(DCent=log2(D)-mean(log2(D)))

#create the model
modS8a <- lme(log2(mutation_rate)~1 + genotype, data=dS8a, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~lower_bound)))

#use model to create values for the lines in the figure
dS8a$pred <- predict(modS8a,level=0)

#create the figure
pS8a <- ggplot(aes(x = D,y = mutation_rate,  colour = genotype, shape = marker), data=dS8a)

pdf("S8A_Fig.pdf",width=4, height=4.4)

pS8a + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), limits=c(5,95), breaks=c(6,30,90))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(7e7,2e8,6e8))) +
  scale_shape_manual(values=c(19,19), guide=FALSE) +
  scale_color_manual(values=c("#246C7F","#4DCDFF"),guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()

#### S8B Fig; first create Model S-XI
# read in the data and create necessary composite columns
dS8b <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4b") %>%
  mutate(DCent=log2(D)-mean(log2(D)))

#create the model
modS8b <- lme(log2(mutation_rate)~DCent*genotype, data=dS8b, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~upper_bound)))

#use model to create values for the lines in the figure
dS8b$pred <- predict(modS8b,level=0)

#create the figure
pS8b <- ggplot(aes(x = D,y = mutation_rate,  colour = genotype, shape = marker), data=dS8b)

pdf("S8B_Fig.pdf",width=4, height=4.4)

pS8b + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), limits=c(5,95), breaks=c(6,30,90))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(8e7,1.5e8,3e8))) +
  scale_shape_manual(values=c(17,17), guide=FALSE) +
  scale_color_manual(values=c("#251DCC","#6BD693"),guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(plot.margin=unit(c(5.5, 15, 5.5, 5.5),"points")) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1),
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()

#### S8C Fig; first create Model S-XIII
# read in the data and create necessary composite columns
dS8c <-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  subset(figure == "4c") %>%
  mutate(CCent=log2(CC)-mean(log2(CC)))

#create the model
modS8c <- lme(log2(mutation_rate)~1, data=dS8c, random=~1|block/plate_ID, weights=varComb(varPower(form = ~fitted(.)), varPower(form = ~upper_bound)))

#use model to create values for the lines in the figure
dS8c$pred <- predict(modS8c,level=0)

#create the figure
pS8c <- ggplot(aes(x = CC,y = mutation_rate,  colour = genotype, shape = included), data=dS8c)

pdf("S8C_Fig.pdf",width=4, height=4.4)

pS8c + 
  
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), breaks=c(500,3000, 12000))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(2e6,1e7,1e8))) +
  scale_shape_manual(values=c(0,15), guide=FALSE) +
  scale_color_manual(values=c("#FF671B","#663300"),guide=FALSE) +
  geom_line(aes(y = 2^pred), lwd=1, colour="#FF671B", alpha=0.8) +
  geom_line(aes(y = 2^pred), lwd=1, colour="#663300", alpha=0.8, lty=2) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()

#### S8D Fig; first create Model S-XV
# read in the data and create necessary composite columns
dS8d <-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  filter(figure %in% c("4d")) %>%
  subset(block == "BB72")  %>%
  mutate(CCent=log2(CC)-mean(log2(CC)))

#create the model
modS8d <- lme(log2(mutation_rate)~CCent, data=dS8d, random=~1|block/plate_ID, weights=varPower(form=~SD_Nt*fitted(.)))

#use model to create values for the lines in the figure
dS8d$pred <- predict(modS8d,level=0)

#create the figure
pS8d <- ggplot(aes(x = CC,y = mutation_rate,  colour = genotype, shape = marker), data=dS8d)

pdf("S8D_Fig.pdf",width=4, height=4.4)

pS8d + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), breaks=c(4,8,12))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(7e7,1e8,2e8))) +
  scale_shape_manual(values=c(18), guide=FALSE) +
  scale_color_manual(values=c("#F2625A"),guide=FALSE) +
  geom_line(aes(group=genotype, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()

#### S9 Fig; first create Model S-XVI
# read in the data and create necessary composite columns
dS9 <- read.csv(file = "S1_Data.csv", header = TRUE) %>%
  filter(!is.na(Ne))  %>%
  mutate(genotype_marker=factor(paste(genotype, marker, sep = '_')))
dS9$DAMP <- factor(dS9$genotype)
levels(dS9$DAMP) <- list("non-plastic" = c("mutT1","mutT2", "PAO1", "PCD1_by", "PCD1_sigma"), 
                         "slightly"   = c("BY4742", "mutH", "mutL", "mutS", "Sigma_1278b"),
                         "plastic" = c("dam", "dinB", "MG1655", "MLH1_sigma", "mutM", "mutY", "S288C", "nei", "metI"))

#create the model
modS9 <- lme(log2(mutation_rate) ~ log2(Ne)*genotype_marker, data=dS9, random=~1|block/plate_ID, weights=varComb(varIdent(form = ~1 | genotype), varPower(form = ~D)))

#use model to create values for the lines in the figure
dS9$pred <- predict(modS9,level=0)

#create the figure
pS9 <- ggplot(aes(x = Ne,y = mutation_rate,  colour = genotype, shape = marker), data=dS9)

pdf("S9_Fig.pdf",width=13, height=7)

pS9 + 
  
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), breaks=c(0.5, 50, 500, 1000))  + 
  scale_x_log10(expression(paste("Effective population size (Ne)")), (breaks=c(3e3, 3e4, 3e5))) +
  facet_grid(~DAMP) +
  geom_line(aes(group=genotype_marker, y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=3) +
  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), 
        panel.background = element_rect(fill = "white"), panel.border = element_blank(),axis.line.x = element_line(color = 'black', size=1), 
        axis.line.y = element_line(color = 'black', size=1), axis.title.x = element_text(face="bold", size=14), 
        axis.title.y = element_text(face="bold", size=14), axis.ticks=element_line(color = 'black', size=2), 
        axis.text.x=element_text(color = 'black', size=14), 
        axis.text.y=element_text(color = 'black', size=14) ) 
graphics.off()

#### S10 Fig; first create Model S-XVII
# read in the data and create necessary composite columns
dS10 <- read.csv(file = "S1_Data.csv", header = TRUE)  %>%
  mutate(m_per_Vt2 = m/(culture_volume*generations), 
         genotype_marker=factor(paste(genotype, marker, sep = '_'))) %>%
  filter(!is.na(m_per_Vt2))
dS10$DAMP <- factor(dS10$genotype)
levels(dS10$DAMP) <- list("non-plastic" = c("mutT1","mutT2", "PAO1", "PCD1_by", "PCD1_sigma"), 
                          "slightly"   = c("BY4742", "mutH", "mutL", "mutS", "Sigma_1278b"),
                          "plastic" = c("dam", "dinB", "MG1655", "MLH1_sigma", "mutM", "mutY", "S288C", "nei", "metI"))

#create the model
modS10 <- lme(log2(m_per_Vt2)~log2(Nt)*genotype_marker, data=dS10, random=~1|block/plate_ID, weights=varPower(form=~SD_m|marker), control=lmeControl(maxIter=500))

#use model to create values for the lines in the figure
dS10$pred <- predict(modS10,level=0)

#create the figure
pS10 <- ggplot(aes(x = Nt,y = m_per_Vt2,  colour = genotype, shape = marker), data=dS10)

pdf("S10_Fig.pdf",width=13, height=7)

pS10 + 
  
  scale_y_log10(expression(paste("Mutational events per ml per division time")), breaks=c(0.01, 0.5, 50))  + 
  scale_x_log10(expression(paste("Population size (Nt)")), (breaks=c(3e5, 3e7, 3e9))) +
  facet_grid(~DAMP) +
  geom_line(aes(y = 2^pred), alpha=0.8, lwd=1)  +
  geom_point(alpha=0.8, size=3) +
  geom_line(colour ="black",lty=2, data=data.frame(Nt=c(3e5,3e9), m_per_Vt2=c(0.005, 50), genotype=c("dam", "dam"), marker=c("rifampicin50", "rifampicin50"))) +
  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), 
        panel.background = element_rect(fill = "white"), panel.border = element_blank(),axis.line.x = element_line(color = 'black', size=1), 
        axis.line.y = element_line(color = 'black', size=1), axis.title.x = element_text(face="bold", size=14), 
        axis.title.y = element_text(face="bold", size=14), axis.ticks=element_line(color = 'black', size=2), 
        axis.text.x=element_text(color = 'black', size=14), 
        axis.text.y=element_text(color = 'black', size=14) ) 
graphics.off()

#### S11 Fig; first create Model S-XVIII
# read in the data and create necessary composite columns
dS11<-read.csv(file = "S1_Data.csv", header = TRUE) %>%
  filter(figure %in% c("S11")) %>%
  mutate(DCent=log2(D)-mean(log2(D)))

#create the model
modS11 <- gls(log2(mutation_rate)~DCent + host + environment, data=dS11)

#use model to create values for the lines in the figure
dS11$pred <- predict(modS11,level=0)

#create the figure
pS11 <- ggplot(aes(x = D,y = mutation_rate,  colour = host, shape = environment), data=dS11)

pdf("S11_Fig.pdf",width=11, height=7)

pS11 + 
  scale_y_log10(expression(paste("Mutation rate (10"^-9, "generation"^-1,")")), (breaks=c(3000,15000,40000)))  + 
  scale_x_log10(expression(paste("Population density (ml"^-1,")")), (breaks=c(8e4, 3e5, 8e5))) +
  geom_line(aes(y = 2^pred), alpha=0.8, lwd=1) +
  geom_point(alpha=0.8, size=4) +
  theme_classic(base_size = 14) +
  theme(axis.line.x= element_line(size=1), 
        axis.line.y = element_line(size=1), 
        axis.ticks=element_line(size=2), 
        axis.text.x=element_text(size=14), 
        axis.text.y=element_text(size=14)) 
graphics.off()