R Notebook - Assembly metrics

#install.packages('googlesheets4')
library(googlesheets4)
library(tidyverse)
library(ggstance)
library(ggbeeswarm)

# make some helpful shortcuts
x.theme.axis.rotate <- theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

standard.textsize <- 10
text.size.within <- (5/14)*(standard.textsize-4)
panel.lab.size <- 10

theme.text.size <- theme(text = element_text(size = standard.textsize))

/ Import googlesheet

gs4_deauth()
#CostaRica_assembly.stats <- read_sheet("https://docs.google.com/spreadsheets/d/1CEDScAVSXuboNu-K_-lZzkPnk9woO_3MFRZ8zgWG3c0/edit?usp=sharing", sheet="Instructor_Testing")

CostaRica_assembly.stats <- read_sheet("https://docs.google.com/spreadsheets/d/1CEDScAVSXuboNu-K_-lZzkPnk9woO_3MFRZ8zgWG3c0/edit#gid=0", sheet="Student_Results")

✔ Reading from WCS_Costa_Rica_2024__Assembly_methods_comparisons.
✔ Range ''Student_Results''.
New names:
• `` -> `...13`

CostaRica_assembly.stats <- CostaRica_assembly.stats %>%
  filter(!is.na(`Student Name`))

Create a variable ‘method_full’ that summarises the methods used

CostaRica_assembly.stats <- CostaRica_assembly.stats %>%
  mutate(method_full=paste0(`Assembler (Unicycler/Dragonflye)`, "__", `Assembly Type (short/long/hybrid)`, "__", `Number of long reads (if used)`)) %>%
  mutate(student_method=paste0(`Student Name`, "_", method_full))

Do some basic plotting

CostaRica_assembly.stats %>%
  arrange(desc(`# contigs`)) %>%
  mutate(method_full=factor(method_full, levels=unique(method_full))) %>%
  ggplot(aes(y=method_full, x=`# contigs`, color=`Assembly Type (short/long/hybrid)`)) +
  geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  theme_bw() + theme(legend.position = 'none') + 
  theme.text.size + 
  labs(y="Method", x="# Contigs", title="# Contigs for each method")

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Assembly.Methods.by.Contig.count.png"), units='mm', width=120, height=70, device='png', dpi=300)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

CostaRica_assembly.stats %>%
  arrange((`N50`)) %>%
  mutate(method_full=factor(method_full, levels=unique(method_full))) %>%
  ggplot(aes(y=method_full, x=`N50`, color=`Assembly Type (short/long/hybrid)`)) +
  geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  theme_bw() + theme(legend.position = 'none') + 
  theme.text.size +
  labs(y="Method", x="N50", title="N50 for each method")

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Assembly.Methods.by.N50.png"), units='mm', width=120, height=70, device='png', dpi=300)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

CostaRica_assembly.stats %>%
  arrange(desc(`# misassemblies`)) %>%
  mutate(method_full=factor(method_full, levels=unique(method_full))) %>%
  ggplot(aes(y=method_full, x=`# misassemblies`, color=`Assembly Type (short/long/hybrid)`)) +
  geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  theme_bw() + theme(legend.position = 'none') + 
  theme.text.size +
  labs(y="Method", x="# Misassemblies", title="# Misassemblies for each method")

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Assembly.Methods.by.Missassembly.count.png"), units='mm', width=120, height=70, device='png', dpi=300)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

Do some more analyses by technique

CostaRica_assembly.stats %>%
  arrange((`N50`)) %>%
  mutate(method_full=factor(`Assembly Type (short/long/hybrid)`, levels=unique(`Assembly Type (short/long/hybrid)`))) %>%
  ggplot(aes(y=method_full, x=`N50`, color=`Assembler (Unicycler/Dragonflye)`)) +
  geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  theme_bw() + theme(legend.position = 'bottom') + 
  theme.text.size +
  labs(y="Assembly Type", x="N50", title="N50 for each Assembly Type")

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Assembly.Type.by.N50.png"), units='mm', width=120, height=60, device='png', dpi=300)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

CostaRica_assembly.stats %>%
  arrange((`N50`)) %>%
  mutate(method_full=factor(`Assembly Type (short/long/hybrid)`, levels=unique(`Assembly Type (short/long/hybrid)`))) %>%
  filter(!is.na(N50)) %>%
  ggplot(aes(y=method_full, x=`N50`, color=factor(`Number of long reads (if used)`))) +
  #ggplot(aes(y=method_full, x=`N50`)) +
  geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  theme_bw() + theme(legend.position = 'bottom') + 
  theme.text.size + 
  facet_grid(`Assembler (Unicycler/Dragonflye)`~.) +
  labs(y="Assembly Type", x="N50", title="N50 for each Assembly Type", color="# Long Reads")

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Assembly.Type.+.readinput.by.N50.png"), units='mm', width=120, height=90, device='png', dpi=300)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

Plot N50 by person

CostaRica_assembly.stats %>%
  arrange((`N50`)) %>%
  #filter(method_full=="Unicycler__hybrid__10000") %>%
  mutate(`student_method`=factor(`student_method`, levels=unique(`student_method`))) %>%
  ggplot(aes(y=`student_method`, x=`N50`, color=`Student Name`)) +
  geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  #geom_barh(width=0.5, stat='identity') + 
  theme_bw() + theme(legend.position = 'none') + 
  theme.text.size + 
  geom_text(aes((N50+500000), `student_method`,label=`Student Name`), size=2.5) +
  labs(y="Student", x="N50", title="N50 for each Student")

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Student.by.N50.png"), units='mm', width=110, height=140, device='png', dpi=900)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

CostaRica_assembly.stats %>%
  arrange(desc(`# contigs`)) %>%
  #filter(method_full=="Unicycler__hybrid__10000") %>%
  mutate(student_method=factor(`student_method`, levels=unique(`student_method`))) %>%
  ggplot(aes(y=`student_method`, x=`# contigs`, color=`Student Name`)) +
  #geom_quasirandom(width=0.4, alpha=0.75, groupOnX=F) +
  geom_quasirandom() +
  #geom_barh(width=0.5, stat='identity') + 
  theme_bw() + theme(legend.position = 'none') + 
  theme.text.size + 
  geom_text(aes((`# contigs`+80), `student_method`,label=`Student Name`), size=2.5) +
  labs(y="Student", x="# Contigs", title="# Contigs for each Student") +
  #coord_cartesian(xlim=c(1,500)) #+ scale_x_log10() +
NULL

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`

ggsave(paste0("Student.by.Contig.count.png"), units='mm', width=110, height=140, device='png', dpi=900)

Orientation inferred to be along y-axis; override with
`position_quasirandom(orientation = 'x')`