Growth_Curve.Rmd

---
title: "Growth_Curve"
author: "ikang"
date: '2020 12 3 '
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Drawing growth curves using R

First, you have to set working directory, and then load packages "openxlsx" and "tidyverse".

```{r, message = FALSE}
setwd("D:/Kin/Protocols/R/R_GrowthCurve2/R_GrowthCurve")
library(openxlsx)
library(tidyverse)
```

## Growth curve without error bars (single tube per each strain)

First, read and save the data from a specific sheet of Excel file.
```{r}
single <- read.xlsx("Growth.xlsx", sheet="Single")
```

Then, transform the data into tidy form.
```{r}
single.tidy <- gather(data=single, key=Days, value = Density, -Strain)
```

You can read and transform together by piping with "%>%".  
If you use piping, the output of the previous command is automatically the input of the next command.
```{r}
single.tidy <- read.xlsx("Growth.xlsx", sheet="Single") %>%
  gather(key=Days, value = Density, -Strain)
```

Now, plot the data.
```{r}
ggplot(single.tidy, aes(x=Days, y=Density, group=Strain, colour=Strain)) +
  geom_line(size=1) + geom_point(size=4)
```

Y-axis of growth curvers usually follows log scale.
```{r}
ggplot(single.tidy, aes(x=Days, y=Density, group=Strain, colour=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10()
```

"Facet" : One plot for one strain
```{r}
ggplot(single.tidy, aes(x=Days, y=Density, group=Strain, color=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() +
  facet_wrap(~Strain)
```


## Growth curve with error bars (triplicate per each strain)
Read the data and make tidy.
```{r}
triple <- read.xlsx("Growth.xlsx", sheet="Triplicate")
triple <- gather(data=triple, key=Days, value = Density, -Strain, -Flask)
```

Grouping the data by Strain AND Days: In this case, 3 flasks are grouped per each strain per each day.
```{r}
triple.gr <- group_by(triple, Strain, Days)
```

Check the result of "grouping".
```{r}
triple.gr
```

Calculate and save summary stats based on the grouping (i.e. per each strain per each day)
```{r}
triple.sum <- summarise(triple.gr, repli=n(), avg=mean(Density), sd=sd(Density))
```
In the above,
number of replicates -> saved into 'repli' column
mean -> saved into 'avg' column
sd (standard deviation) -> saved into 'sd' column

All the above steps can be piped as below.
```{r}
triple.sum <- read.xlsx("Growth.xlsx", sheet="Triplicate") %>%
  gather(key=Days, value = Density, -Strain, -Flask, convert=T) %>%
  group_by(Strain, Days) %>%
  summarise(repli=n(), avg=mean(Density), sd=sd(Density), se=sd(Density)/sqrt(repli))
```

Check the result.
```{r}
triple.sum
```

Plot the result.
First, without error bar
```{r}
ggplot(triple.sum, aes(x=Days, y=avg, group=Strain, colour=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() + 
  scale_x_continuous(breaks=seq(0,10,by=1))
```

Then, with error bar
```{r}
ggplot(triple.sum, aes(x=Days, y=avg, group=Strain, colour=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() +
  geom_errorbar(aes(ymin=avg-sd, ymax=avg+sd), width=0.1, size=0.5)
```

Refine the Y-axis label
```{r}
ggplot(triple.sum, aes(x=Days, y=avg, group=Strain, colour=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() +
  geom_errorbar(aes(ymin=avg-sd, ymax=avg+sd), width=0.1, size=0.5) +
  ylab('Cells '~mL^-1)
```

Facet the graph
```{r}
ggplot(triple.sum, aes(x=Days, y=avg, group=Strain, color=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() +
  geom_errorbar(aes(ymin=avg-sd, ymax=avg+sd), width=0.1, size=0.5) +
  ylab('Cells '~mL^-1) +
  facet_wrap(~Strain)
```

Use error ribbon instead of error bar.
```{r}
ggplot(triple.sum, aes(x=Days, y=avg, group=Strain, color=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() +
  geom_ribbon(aes(ymin = avg-sd, ymax = avg+sd, fill=Strain), alpha=0.3, linetype=0) +
  ylab('Cells '~mL^-1)
```

Facet with both error bar and error ribbon
```{r}
ggplot(triple.sum, aes(x=Days, y=avg, group=Strain, color=Strain)) +
  geom_line(size=1) + geom_point(size=4) + 
  scale_y_log10() +
  geom_errorbar(aes(ymin=avg-sd, ymax=avg+sd), width=0.1, size=0.5) +
  geom_ribbon(aes(ymin = avg-sd, ymax = avg+sd, fill=Strain), alpha=0.3, linetype=0) +
  ylab('Cells '~mL^-1) +
  facet_wrap(~Strain)
```