Lambda conversion rates

tl;dr: It looks like we have ~ 98.6% conversion efficiency for CpG sites.

I wanted to look at the Methylation conversion rates for the lambda spike in for our samples. This is used as a proxy for sample methylation conversion rate because lambda methylation rate is usually known. We goofed up a bit and used a potentially methylated lambda strain. Oops.

First I read in the data I curated from a bunch of bismark files.

library(readr)
lambda_conversion_rates <- read_csv("~/Documents/Genewiz_hdd/lambda_conversion_rates.csv")

Parsed with column specification:
cols(
  Sample = col_character(),
  `Mapping Eff` = col_double(),
  `CpG Methylation Rate` = col_double(),
  `CHG Methylation Rate` = col_double(),
  `CHH Methylation Rate` = col_double()
)

Here I plot histograms and boxplots (with a stripchart of the actual data overlayed) for Mapping Efficiency, as well as mean and standard deviation. P

hist(lambda_conversion_rates$`Mapping Eff`)

boxplot(lambda_conversion_rates$`Mapping Eff`, outline = FALSE, ylim = c(min(lambda_conversion_rates$`Mapping Eff`) - 0.1, max(lambda_conversion_rates$`Mapping Eff`) + 0.1))
stripchart(lambda_conversion_rates$`Mapping Eff`, add = TRUE, axes = F, vertical = T, jitter = 0.02, method = "jitter", pch = 16)

mean(lambda_conversion_rates$`Mapping Eff`)

[1] 0.5692308

sd(lambda_conversion_rates$`Mapping Eff`)

[1] 0.2790309

Next I look at CpG Methlaytion rate in the same way

hist(lambda_conversion_rates$`CpG Methylation Rate`)

boxplot(lambda_conversion_rates$`CpG Methylation Rate`, outline = FALSE, ylim = c(1.1, 2.3))
stripchart(lambda_conversion_rates$`CpG Methylation Rate`, add = TRUE, axes = F, vertical = T, jitter = 0.02, method = "jitter", pch = 16)

mean(lambda_conversion_rates$`CpG Methylation Rate`)

[1] 1.419231

sd(lambda_conversion_rates$`CpG Methylation Rate`)

[1] 0.1836861

And next CHG methylation rates.

hist(lambda_conversion_rates$`CHG Methylation Rate`)

boxplot(lambda_conversion_rates$`CHG Methylation Rate`, outline = FALSE, ylim = c(min(lambda_conversion_rates$`CHG Methylation Rate`) - 0.1, max(lambda_conversion_rates$`CHG Methylation Rate`) + 0.1))
stripchart(lambda_conversion_rates$`CHG Methylation Rate`, add = TRUE, axes = F, vertical = T, jitter = 0.02, method = "jitter", pch = 16)

mean(lambda_conversion_rates$`CHG Methylation Rate`)

[1] 4.296154

sd(lambda_conversion_rates$`CHG Methylation Rate`)

[1] 0.4818109

Then CHH methylation rates the same way.

hist(lambda_conversion_rates$`CHH Methylation Rate`)

boxplot(lambda_conversion_rates$`CHH Methylation Rate`, outline = FALSE, ylim = c(min(lambda_conversion_rates$`CHH Methylation Rate`) - 0.1, max(lambda_conversion_rates$`CHH Methylation Rate`) + 0.1))
stripchart(lambda_conversion_rates$`CHH Methylation Rate`, add = TRUE, axes = F, vertical = T, jitter = 0.02, method = "jitter", pch = 16)

mean(lambda_conversion_rates$`CHH Methylation Rate`)

[1] 1.073077

sd(lambda_conversion_rates$`CHH Methylation Rate`)

[1] 0.2010905

Add a new chunk by clicking the Insert Chunk button on the toolbar or by pressing Ctrl+Alt+I.

When you save the notebook, an HTML file containing the code and output will be saved alongside it (click the Preview button or press Ctrl+Shift+K to preview the HTML file).