1 Data Import and Reshaping

rawdata <- read.csv("data_input/Example01.csv")

The initial data are in wide format, we need them in long format. For this transformation we can use reshape:

longdata <- reshape(rawdata, varying = list(c("Bio.1", "Bio.2", "Bio.3", "Bio.4")),
    v.names = "Biofilm", timevar = "Experiment", direction = "long")
longdata$id <- factor(longdata$id, labels = c("WT_TSB", "MutA_TSB", "MutB_TSB", "MutAB_TSB",
    "WT_CDM", "MutA_CDM", "MutB_CDM", "MutAB_CDM"))

2 Basic Plots

2.1 Scatterplots

ggplot(longdata)

ggplot(longdata, aes(x = Strain, y = Biofilm)) +
  geom_point()

ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_point()

ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_boxplot()

ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_boxplot() +
  geom_point(position = position_dodge(0.75), alpha=0.5)

ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_point(alpha=0.5) +
  stat_summary(fun = "mean", geom = "point", size = 4)

ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  stat_summary(fun = "mean", geom = "point", size = 4)

2.2 Bar plots

ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  stat_summary(fun = "mean", geom = "bar")

ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge())

ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge()) +
  stat_summary(fun.data = "median_hilow", geom = "errorbar", position=position_dodge(.9))

ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun.data = "median_hilow", geom = "errorbar", position=position_dodge(.9), width = 0.2) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge())

fun.data = “median_hilow” (When the default conf.int=0.95 is used, the lower and upper quantiles computed are 0.025 and 0.975.) See also: “mean_cl_boot” / “mean_sdl”

3 Styling plots

ugly.plot <- ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun.data = "median_hilow", geom = "errorbar", position=position_dodge(.9), width = 0.2) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge())

ugly.plot +
  theme_classic()

ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40"))

ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) +
  scale_y_continuous(expand = c(0, 0))

ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) +
  scale_y_continuous(expand = c(0, 0)) +
  ylab("Biofilm(OD600)") + xlab("")

4 Facets

ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) +
  scale_y_continuous(expand = c(0, 0)) +
  ylab("Biofilm(OD600)") + xlab("") +
  facet_wrap(~ Media)

5 Statistics

5.1 One-way ANOVA

ANOVA <- aov(Biofilm ~ id, longdata)

##             Df Sum Sq Mean Sq F value Pr(>F)    
## id           7 0.3844 0.05492   440.9 <2e-16 ***
## Residuals   24 0.0030 0.00012                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

RESULT: There is a statistically significant difference between these conditions, with a p-value of 9.2101939^{-24}

5.2 Two-way ANOVA

ANOVA2 <- aov(Biofilm ~ Strain + Media, longdata)

##             Df  Sum Sq Mean Sq F value   Pr(>F)    
## Strain       3 0.25319 0.08440   49.07 4.62e-11 ***
## Media        1 0.08778 0.08778   51.03 1.11e-07 ***
## Residuals   27 0.04644 0.00172                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

RESULT: There is a statistically significant difference between strains, with a p-value of 4.619916^{-11}. There is a statistically significant difference between media, with a p-value of 1.1074002^{-7}.

5.3 T-test Pairwise comparisons

Pairwise comparisons with Bonferroni correction

BONF <- pairwise.t.test(longdata$Biofilm,longdata$id, p.adj = "bonf")
BONF.table <- xtable(BONF$p.value, digits = 4)
print(BONF.table, type = "html")

	WT_TSB	MutA_TSB	MutB_TSB	MutAB_TSB	WT_CDM	MutA_CDM	MutB_CDM
MutA_TSB	0.0000
MutB_TSB	0.0732	0.0000
MutAB_TSB	0.0000	0.0000	0.0000
WT_CDM	0.0000	0.0002	0.0000	0.0000
MutA_CDM	0.0000	0.0000	0.0000	0.0001	0.0000
MutB_CDM	0.0000	0.0000	0.0000	0.0001	0.0000	1.0000
MutAB_CDM	0.0000	0.0000	0.0000	1.0000	0.0000	0.0000	0.0000

Pairwise comparisons with Tukey correction

TUKEY <- TukeyHSD(ANOVA, "id")
TUKEY.table <- xtable(TUKEY$id, digits = 4)
print(TUKEY.table, type = "html")

	diff	lwr	upr	p adj
MutA_TSB-WT_TSB	-0.1422	-0.1684	-0.1161	0.0000
MutB_TSB-WT_TSB	-0.0265	-0.0526	-0.0004	0.0453
MutAB_TSB-WT_TSB	-0.2895	-0.3156	-0.2634	0.0000
WT_CDM-WT_TSB	-0.0970	-0.1231	-0.0709	0.0000
MutA_CDM-WT_TSB	-0.2430	-0.2691	-0.2169	0.0000
MutB_CDM-WT_TSB	-0.2410	-0.2671	-0.2149	0.0000
MutAB_CDM-WT_TSB	-0.2962	-0.3224	-0.2701	0.0000
MutB_TSB-MutA_TSB	0.1157	0.0896	0.1419	0.0000
MutAB_TSB-MutA_TSB	-0.1472	-0.1734	-0.1211	0.0000
WT_CDM-MutA_TSB	0.0453	0.0191	0.0714	0.0002
MutA_CDM-MutA_TSB	-0.1008	-0.1269	-0.0746	0.0000
MutB_CDM-MutA_TSB	-0.0988	-0.1249	-0.0726	0.0000
MutAB_CDM-MutA_TSB	-0.1540	-0.1801	-0.1279	0.0000
MutAB_TSB-MutB_TSB	-0.2630	-0.2891	-0.2369	0.0000
WT_CDM-MutB_TSB	-0.0705	-0.0966	-0.0444	0.0000
MutA_CDM-MutB_TSB	-0.2165	-0.2426	-0.1904	0.0000
MutB_CDM-MutB_TSB	-0.2145	-0.2406	-0.1884	0.0000
MutAB_CDM-MutB_TSB	-0.2697	-0.2959	-0.2436	0.0000
WT_CDM-MutAB_TSB	0.1925	0.1664	0.2186	0.0000
MutA_CDM-MutAB_TSB	0.0465	0.0204	0.0726	0.0001
MutB_CDM-MutAB_TSB	0.0485	0.0224	0.0746	0.0001
MutAB_CDM-MutAB_TSB	-0.0068	-0.0329	0.0194	0.9874
MutA_CDM-WT_CDM	-0.1460	-0.1721	-0.1199	0.0000
MutB_CDM-WT_CDM	-0.1440	-0.1701	-0.1179	0.0000
MutAB_CDM-WT_CDM	-0.1992	-0.2254	-0.1731	0.0000
MutB_CDM-MutA_CDM	0.0020	-0.0241	0.0281	1.0000
MutAB_CDM-MutA_CDM	-0.0532	-0.0794	-0.0271	0.0000
MutAB_CDM-MutB_CDM	-0.0552	-0.0814	-0.0291	0.0000

Basic ggplot2