Install

Install dependent R packages

On Windows, you should have Rtools.

install.packages("dplyr")
install.packages("tidyr")
install.packages("")
install.packages("ggpubr")

library(BiocManager)
BiocManager::install("limma")

library(remotes)
remotes::install_github(repo="jdreyf/ezlimma", build_opts = c("--no-resave-data", "--no-manual"), build_vignettes = FALSE)

Usage

Example run

######################################################################
# packages
######################################################################
library(tidyRnaEdit)

######################################################################
# Read-in example data
######################################################################
# Read-in the example dataset.
data(T15H10,package="tidyRnaEdit")

# This example dataset contains two files. 
# 1. editing_rate matrix: merged_data
# 2. sample_info: sample_info

# get a look at this dataset
merged_data %>% head
sample_info %>% head

# Read-in annotation file.
data(anno_hg38_REDIportal,package="tidyRnaEdit")

# get a look at annotation infomation.
anno %>% head

######################################################################
# Count num. of editing types
######################################################################
# count
edit_type_count = edit_type_count(merged_data)
# save
write.csv(edit_type_count, file=paste0(path_save_type,"/mat_raw_count.csv"), row.names = FALSE)

# plot without save
plot_edit_type_count(edit_type_count)

# plot and save
path_save_type="."  # change to your save_directory_path
plot_edit_type_count(edit_type_count, save = TRUE,
                     file=paste0(path_save_type,"/plot_edit_type_count.pdf"))

######################################################################
# Filtering variants
######################################################################
mat_qc = filter_variant(merged_data, sample_info,
                        edit_rate_indi = 0.01,
                        edit_rate_mean = 0.05,
                        missing_rate = NA,
                        left_n_total = NA,
                        left_n_in_each_group = 5)

path_save="."  # change to your save_directory_path
write.csv(mat_qc, file=gzfile(paste0(path_save,"/mat_qc.csv.gz")), row.names = FALSE)

mat_qc[duplicated(mat_qc$symbol) == TRUE,]

######################################################################
# diff_analy_limma for genes
######################################################################
mat_mean = calc_mean(mat_qc, sample_info = sample_info)

diff_res = diff_edit_site(mat_qc,
                          sample_info = sample_info,
                          comparison = c("Han", "Tibetan"))  # control_group = "Han"

# threshold
FDR_thres = 0.05
deltaEdit_thres = 0.30

# filter sig symbol
diff_sig = diff_res %>%
  dplyr::filter(abs(`Tibetan-Han.delta`)> deltaEdit_thres) %>%
  dplyr::filter(`Tibetan-Han.FDR` < FDR_thres)

# # diff_summary
# TODO
# omic6_diff_limma_num = meth_diff_limma_num(omic6_diff, deltaB_thres = deltaEdit_thres)
# omic6_diff_limma_num_total = meth_diff_limma_num_total(omic6_diff, deltaB_thres = deltaEdit_thres)

######################################################################
# annotate
######################################################################
diff_sig_anno = anno_edit_site(diff_sig, anno)

path_save="."  # change to your save_directory_path
write.csv(diff_sig_anno, file=paste0(path_save,"/diff_sig_anno.csv"), row.names = FALSE)

# extract recoding site
diff_sig_recoding_site = diff_sig_anno %>% dplyr::filter(Func.wgEncodeGencodeBasicV34=="exonic")

diff_sig_recoding_site %>% head

Run with your data

Download the example data of REDItools outTables.

We provided an example data at github, which you can download at your PC.

Then, unzip the tidyRnaEdit_exampledata.zip, which are,

The files named *outTable.gz are REDItools outTables. Only chr17 is provided for test.
The file reditools_sample.xlsx is a dataframe of 3 columns, including filepath, sampleid, group.

NOTE:

*outTable.gz is the standard output by REDItools.
When you generate your reditools_sample.xlsx file, You should KEEP (not modify) the column names.

Download the annotation file. (if using hg38, pass this step)

Now, tidyRnaEdit has already included the annotation file on human (hg38). You can easily use it without downloading.

# just type
library(tidyRnaEdit)
data(anno_hg38_REDIportal,package="tidyRnaEdit")

# get a look at this annotation file.
anno %>% head

# The annotation file on human (hg38) is already to use in this package, namely, the `anno`.

For other genome version of human and other species, you should download elsewhere.

We recommend a well-curated annotated database, which is available at REDIportal.

Tutorials with the example data.

######################################################################
# packages
######################################################################
library(tidyRnaEdit)


######################################################################
# Path
######################################################################
#--------- work space ------------
setwd("/data/pingj/soft/tidyRnaEdit_exampledata/")  # change to your_path_where_saved_the_outTables

#--------- path_data ------------
# sample_info (3 cols: filename, sampleid, group)
path_sample = "reditools_samples.xlsx"  # change to your_path_where_saved_the_outTables

# anno (using `anno` in our package, or you can download from REDIportal)
path_anno = "TABLE1_hg38.txt.gz"  # change to your_path_where_saved_the_outTables

#--------- path_save ------------
# root
path_save = "."  # change to your save_directory_path
# editing_type
path_save_type        = paste0(path_save, "")   # change to your save_directory_path
# OverallEditing
path_save_overalledit = paste0(path_save, "")   # change to your save_directory_path

######################################################################
# Read-in data
######################################################################
# sample info
sample_info = readxl::read_excel(path_sample)

# anno
anno = read.table(gzfile(path_anno), sep ='\t', fill=TRUE, header = TRUE)

# editing data
merged_data = read_redit(sample_info, soft = "reditools_known")

# save
write.csv(merged_data, file=gzfile(paste0(path_save,"/mat_raw.csv.gz")), row.names = FALSE)

merged_data = read.csv(gzfile(paste0(path_save,"/mat_raw.csv.gz")),header=TRUE)

# Then, you can do down-stream analysis, follow the `example run` section. Enjoy!

JPingAMMS / tidyRnaEdit