The goal of methcon5 is to identify and rank CpG DNA methylation conservation along the human genome. Specifically, it includes bootstrapping methods to provide ranking which should adjust for the differences in length as without it short regions tend to get higher conservation scores.
The following repository includes an analysis in which this package was used.
Please note that the name of the package is in all lowercase.
You can install the released version of methcon5 from CRAN with:
And the development version from GitHub with:
First we apply the meth_aggregate
function. This will take the columns specified in value
and apply the fun
stratified according to id
. In this case, we want to calculate the mean meth value within each gene.
library(methcon5)
sample_ii <- fake_methylation %>%
meth_aggregate(id = gene, value = meth, fun = mean)
sample_ii
#> # A tibble: 500 x 3
#> gene meth n
#> * <int> <dbl> <int>
#> 1 1 0.509 10
#> 2 2 0.817 6
#> 3 3 0.577 5
#> 4 4 0.279 9
#> 5 5 0.318 5
#> 6 6 0.427 6
#> 7 7 0.736 4
#> 8 8 0.546 2
#> 9 9 0.328 7
#> 10 10 0.202 6
#> # … with 490 more rows
Next we use the meth_bootstrap
function. This will take the summarized data.frame calculated earlier along with the original dataset. The function with return the original data.frame with the new column attached to the end, which makes it ideal for piping to apply different methods to the same data.
adjusted <- sample_ii %>%
meth_bootstrap(reps = 100) %>%
meth_bootstrap(reps = 100, method = "perm_v2") %>%
meth_bootstrap(reps = 100, method = "perm_v3")
adjusted
#> # A tibble: 500 x 6
#> gene meth n meth_perm_v1 meth_perm_v2 meth_perm_v3
#> * <int> <dbl> <int> <dbl> <dbl> <dbl>
#> 1 1 0.509 10 0.52 0.6 0.39
#> 2 2 0.817 6 0 0.01 0
#> 3 3 0.577 5 0.28 0.33 0.42
#> 4 4 0.279 9 1 0.91 0.9
#> 5 5 0.318 5 0.96 0.81 0.76
#> 6 6 0.427 6 0.78 0.69 0.63
#> 7 7 0.736 4 0.05 0.16 0.15
#> 8 8 0.546 2 0.38 0.42 0.54
#> 9 9 0.328 7 1 0.79 0.81
#> 10 10 0.202 6 1 1 0.95
#> # … with 490 more rows
library(ggplot2)
ggplot(adjusted, aes(meth_perm_v1, meth_perm_v2, color = n)) +
geom_point() +
scale_color_viridis_c() +
theme_minimal()
We gratefully acknowledge funding from NIH awards 1P01CA196569 and R21 CA226106.