R Notebook

library(xlsx)
library(plotly)
library(ggthemes)

plotdata = day1_3[1:40,] # do graph also for 50 
plotdata = droplevels(plotdata)
plotdata$gene = factor(plotdata$gene, levels =  c(as.character(plotdata$gene)))
plotdata$AbsPCA_variance = abs(plotdata$PCA_variance)
plotdata

plot1 = plot_ly( data = plotdata
        ,x = ~gene
        ,y = abs(plotdata$PCA_variance)
        ,color = ~regulator
        ,type = 'bar'
)
plot1 = layout(plot1
       ,legend = list(title = "Regulator", y = 1, x = 0.9)
       ,yaxis = list(range = c(min(abs(plotdata$PCA_variance))*0.999, max(abs(plotdata$PCA_variance))*1.001), title = "Absolute PCA variance")
       ,width = 1000
       ,margin = list(b=-20)
       )

Specifying width/height in layout() is now deprecated.
Please specify in ggplotly() or plot_ly()

plot1

n too large, allowed maximum for palette Set2 is 8
Returning the palette you asked for with that many colors
n too large, allowed maximum for palette Set2 is 8
Returning the palette you asked for with that many colors
n too large, allowed maximum for palette Set2 is 8
Returning the palette you asked for with that many colors
n too large, allowed maximum for palette Set2 is 8
Returning the palette you asked for with that many colors

# plotly_IMAGE(plot1, format = "png", out_file = "40.png")
# plot(abs(plotdata$PCA_variance))

plot1 = ggplot(data = plotdata, aes(x = gene, y = AbsPCA_variance, color = regulator, label = regulator)) + 
  geom_point(size = 10) +
  geom_segment(aes(x = gene, xend = gene, 
                   y = 0.0226, yend = AbsPCA_variance, 
                   color = regulator), size = 2, alpha = 0.5) + 
  scale_y_continuous(limits = c(0.0226, 0.02288)) +
  geom_text(color = 'white', size = 3) + 
  theme_minimal() +
  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank()) + 
  coord_flip() +
  labs(y = "Absolute PCA Variance", x = "Genes", color = "Regulator", family = 'TT Times New Roman')  
  # scale_colour_brewer(palette = 'Paired')
  # scale_fill_manual(values=c("#CC6666", "#9999CC", "#66CC99"))
plot1

library(extrafont)

Registering fonts with R

font_import()

library(ggalt)
plot2 = ggplot(data = plotdata, aes(x = plotdata$gene, y = plotdata$AbsPCA_variance, color = plotdata$regulator, label = plotdata$regulator)) + 
  geom_lollipop(size = 5) + 
  # scale_y_continuous(limits = c(0.0227, 0.02288)) +
  theme_minimal()
plot2

# Ilana ask for table like that
write.csv(day1_3[1:50,c(1,10,2,4,5,6)], file = "Data from Tsviya about genes variance.csv")

Ilana manually currated some “unknown”" regulators and find matching regulators for top 50 hits.

additional = read.xlsx("data/genes_with_highestVariance_July18 +manually_currated.xlsx",1)
levels(additional$regulator) = c(levels(additional$regulator), levels(additional$manually.regulator))
additional$regulator[!is.na(additional$manually.regulator)] = additional$manually.regulator[!is.na(additional$manually.regulator)]

plotdata = additional[1:50,] # do graph also for 50 
plotdata = droplevels(plotdata)
plotdata$gene = factor(plotdata$gene, levels =  c(as.character(plotdata$gene)))
plotdata$AbsPCA_variance = abs(plotdata$PCA_variance)
plotdata
plot2 = plot_ly( data = plotdata
        ,x = ~gene
        ,y = abs(plotdata$PCA_variance)
        ,color = ~regulator
        ,type = 'bar'
        ,width = 1200
)
plot2 = layout(plot2
       ,legend = list(title = "Regulator")
       ,yaxis = list(range = c(min(abs(plotdata$PCA_variance))*0.999, max(abs(plotdata$PCA_variance))*1.001))
       ,xaxis = list(labels = "outside", tickfont=list(size = 16), automargin = T, showline = F, zeroline = T)
)
plot2

Source code available here.