04-MAW-PIII.Rmd

---
title: "OPEN & REPRODUCIBLE MICROBIOME DATA ANALYSIS SPRING SCHOOL 2018"
author: "Sudarshan"
date: "`r Sys.Date()`"
output: bookdown::gitbook
site: bookdown::bookdown_site
---

# Composition plots  

Barplots are a one way of visualising the composition of your samples.    

We will use the filtered phyloseq object from **Set-up and Pre-processing** section.  

**Load packages** 

```{r, warning=FALSE, message=FALSE}

library(microbiome) # data analysis and visualisation
library(phyloseq) # also the basis of data object. Data analysis and visualisation
library(RColorBrewer) # nice color options
library(ggpubr) # publication quality figures, based on ggplot2
library(dplyr) # data handling  

```

  
```{r}

ps1 <- readRDS("./phyobjects/ps.ng.tax.rds")

# use print option to see the data saved as phyloseq object.

print(ps1)

```

## Barplot counts 

```{r, fig.height= 6, fig.width=20, warning=FALSE, message= FALSE}

ps1.com <- ps1

# if you have dada2/deblur output and sequences as taxa names, then you can change them as follows
taxa_names(ps1.com) <- paste0("ASV_", rownames(tax_table(ps1.com)))

# We need to set Palette
taxic <- as.data.frame(ps1.com@tax_table) # this will help in setting large color options

# colourCount = length(unique(taxic$Family))  #define number of variable colors based on number of Family (change the level accordingly to phylum/class/order)
# getPalette = colorRampPalette(brewer.pal(12, "Paired"))  # change the palette as well as the number of colors will change according to palette.

taxic$OTU <- rownames(taxic) # Add the OTU ids from OTU table into the taxa table at the end.
colnames(taxic) # You can see that we now have extra taxonomy levels.

taxmat <- as.matrix(taxic) # convert it into a matrix.
new.tax <- tax_table(taxmat) # convert into phyloseq compatible file.
tax_table(ps1.com) <- new.tax # incroporate into phyloseq Object


# now edit the unclassified taxa
tax_table(ps1.com)[tax_table(ps1.com)[, "Family"] == "", "Family"] <- "Unclassified family"

# it would be nice to have the Taxonomic names in italics.
# for that we set this

guide_italics <- guides(fill = guide_legend(label.theme = element_text(
  size = 15,
  face = "italic", colour = "Black", angle = 0
)))


## Now we need to plot at family level, we can do it as follows:

# first remove the phy_tree

ps1.com@phy_tree <- NULL

# Second merge at family level

ps1.com.fam <- microbiome::aggregate_top_taxa(ps1.com, "Family", top = 10)

plot.composition.COuntAbun <- plot_composition(ps1.com.fam) + theme(legend.position = "bottom") +
  scale_fill_brewer("Family", palette = "Paired") + theme_bw() +
  theme(axis.text.x = element_text(angle = 90)) +
  ggtitle("Relative abundance") + guide_italics + theme(legend.title = element_text(size = 18))

plot.composition.COuntAbun
#ggsave("./Test_Outputfiles/Family_barplot_CountAbundance.pdf", height = 6, width = 8)


```

This plot is based on the reads per sample. In the next step, we plot the relative abundance.

## Barplot relative abundance 

Make it relative abundance

```{r, fig.height= 6, fig.width=20, warning=FALSE, message= FALSE}

# the previous pseq object ps1.com.fam is only counts.

# Use traqnsform function of microbiome to convert it to rel abun.
ps1.com.fam.rel <- microbiome::transform(ps1.com.fam, "compositional")

plot.composition.relAbun <- plot_composition(ps1.com.fam.rel,
                                             sample.sort = "scientific_name",
                                             x.label = "env_material") 
plot.composition.relAbun <- plot.composition.relAbun + theme(legend.position = "bottom") 
plot.composition.relAbun <- plot.composition.relAbun + scale_fill_brewer("Family", palette = "Paired") + theme_bw() 
plot.composition.relAbun <- plot.composition.relAbun + theme(axis.text.x = element_text(angle = 90)) 
plot.composition.relAbun <- plot.composition.relAbun + ggtitle("Relative abundance") + guide_italics + theme(legend.title = element_text(size = 18))

print(plot.composition.relAbun)

#ggsave("./figures/Family_barplot_RelAbundance.pdf", height = 6, width = 8)


```

### Barplot customize 

```{r}

data.com <- plot.composition.relAbun$data
colnames(data.com)

```

```{r}


p.com <- ggplot(data.com, aes(x = Sample, y = Abundance, fill = Tax))
p.com <- p.com + geom_bar(position = "stack", stat = "identity")
p.com <- p.com + scale_x_discrete(labels = data.com$xlabel, breaks = data.com$Sample)
p.com <- p.com + facet_grid(~xlabel, scales = "free") + theme_bw()
p.com <- p.com + scale_fill_brewer("Family", palette = "Paired")
p.com <- p.com + rremove("x.text")

ggsave("./figures/Composition plots.pdf", height = 4, width = 6)


```

For more information [Microbiome tutorial](http://microbiome.github.io/microbiome/Composition.html)   

## Heatmaps   

These are a good alternative to barplots (if done right).  

```{r}

# base plot
p.heat <- ggplot(data.com, aes(x = Sample, y = Tax)) + geom_tile(aes(fill = Abundance)) 

# Change color
p.heat <- p.heat + scale_fill_distiller("Abundance", palette = "RdYlBu") + theme_bw() 

# Make bacterial names italics
p.heat <- p.heat + theme(axis.text.y = element_text(colour = 'black', 
                                                    size = 10, 
                                                    face = 'italic')) 
# Make seperate samples based on main varaible
p.heat <- p.heat + facet_grid(~xlabel, 
                              scales = "free") + rremove("x.text") 

p.heat <- p.heat + ylab("Family")

#Clean the x-axis
p.heat <- p.heat + theme(axis.title.x=element_blank(),
                     axis.text.x=element_blank(),
                     axis.ticks.x=element_blank()) 

# Clean the facet label box
p.heat <- p.heat + theme(legend.key = element_blank(), 
                     strip.background = element_rect(colour="black", fill="white"))

print(p.heat)

ggsave("./figures/Heatmap.pdf", height = 4, width = 6)


# + geom_text(aes(label = round(Abundance)), size = 0.4)
```


**Extra**  

Following is an example of customizing the plot using ggpubr.  

```{r}

# we use count data at family level from the barplot for counts
ps_df <- microbiomeutilities::phy_to_ldf(ps1.com.fam, 
                                         transform.counts = "compositional")

colnames(ps_df)
# this data.frame can be used to customize several plots.  

# example boxplot at family level

p.box <- ggstripchart(ps_df, "scientific_name", "Abundance", 
                      facet.by = "Family", color = "body_product",
                      palette = "jco"
                      )

p.box + rremove("x.text")

```


```{r}

sessionInfo()

```