find_tails: Estimates poly(A)/(T) tail lengths in Oxford Nanopore RNA and...

In adnaniazi/tailfindr: Poly(A) Tail Length Estimator for Oxford Nanopore RNA and DNA Reads

Estimates poly(A)/(T) tail lengths in Oxford Nanopore RNA and DNA reads

Description

This function estimates poly(A) tail length in RNA reads, and both poly(A) and poly(T) tail lengths in DNA reads. It can operate on reads base called with any version of Albacore and Guppy using either the standard or the recent 'flip-flop' model. The function outputs a CSV file containing poly(A) tail information organised by the read ID; it also returns the same information as a tibble for further processing by the end-user. Currently, the algorithm works only on ONT 1D reads.

Usage

find_tails(fast5_dir, save_dir, csv_filename = "tails.csv",
  num_cores = 1, basecall_group = "Basecall_1D_000",
  save_plots = FALSE, plot_debug_traces = FALSE,
  plotting_library = "rbokeh", ...)

Arguments

fast5_dir	character string. Full path of the directory to search the basecalled fast5 files in. The Fast5 files can be single or multi-fast5 file. The directory is searched recursively.
save_dir	character string. Full path of the directory where the CSV file containing the tail-length information should be stored. If save_plots is set to TRUE, then a plots directory is also created within the save_dir.
csv_filename	character string ["tails.csv"]. Filename of the CSV file in which to store the tail length data
num_cores	numeric [1]. Num of physical cores to use in processing the data. Always use 1 less than the number of cores at your disposal.
basecall_group	a character string ["Basecall_1D_000"]. Name of the level in the Fast5 file hierarchy from which tailfindr should read the data.
save_plots	logical [FALSE]. If set to TRUE, a plots directory will be created within the save_dir, and plots showing poly(A) and poly(T) tails in the raw squiggle will be saved in this plots directory. Creating plots and saving them to the disk is a slow process. We recommend that you keep this option set to FALSE. If you still want to create plots, we recommend that you run tailfindr on a subset of reads. Plots are automatically named by concatenating read ID with the name of the Fast5 file containing this read; the read ID and fast5 file name are separated by two underscores (__).
plot_debug_traces	logical [FALSE]. This option works only if save_plots option is also set to TRUE.If set to TRUE, debugging information is plotted in the plots as well. This includes mean signal, slope signal,thresholds, smoothened signal, etc. We use this option internally to debug our algorithm.
plotting_library	character string ["rbokeh"]. rbokeh is the default plotting library used if save_plots is set to TRUE. The plots will be saved as HTML files in the /save_dir/plots directory. You can open these HTLM files in any web-browser and interactively view the plots showing the tail region in the raw squiggle. If this option is set to 'ggplot2', then the polts will be saved as static .png files.
...	list. A list of optional parameters. This is currently, reserved for internal use only.

Value

A data tibble containing tail information organzied by the read ID is returned. Always save this returned tibble in a variable (see examples below), otherwise the long tibble will be printed to the console, which may hang up your R session.

A CSV file containing the same information is also saved on disk in the save_dir.

Examples

## Not run: 

library(tailfindr)

# 1. Suppose you have 11 cores at your disposal, then you should run tailfindr
# on your data as following:
df <- find_tails(fast5_dir = system.file('extdata', 'rna', package = 'tailfindr'),
                 save_dir = '~/Downloads',
                 csv_filename = 'rna_tails.csv',
                 num_cores = 10)
# In the above example, we have  used tailfindr on example RNA reads
# present in the tailfindr package. You should substitute the path of
# your data for the fast5_dir parameter.

# 2. If you want to save interactive HTML plots using rbokeh,
# then you should run tailfindr as following:
df <- find_tails(fast5_dir = system.file('extdata', 'cdna', package = 'tailfindr'),
                 save_dir = '~/Downloads',
                 csv_filename = 'cdna_tails.csv',
                 num_cores = 10,
                 save_plots = TRUE,
                 plotting_library = 'rbokeh')

# 3. If you also want to plot debug traces, then you should run tailfindr as
# below:
df <- find_tails(fast5_dir = system.file('extdata', 'cdna', package = 'tailfindr'),
                 save_dir = '~/Downloads',
                 csv_filename = 'cdna_tails.csv',
                 num_cores = 10,
                 save_plots = TRUE,
                 plot_debug_traces = TRUE,
                 plotting_library = 'rbokeh')

# N.B.: Making and saving plots is a computationally slow process.
# Only generate plots by running tailfindr on a small subset of your reads.

# 4. By default, tailfindr uses Events/Move table in the Basecall_1D_000
# section of the FAST5 file. If you want tailfindr to pick Events/Move table
# from some other section of the FAST5 file -- lets say Basecall_1D_001--
# then you should use tailfindr like below:
df <- find_tails(fast5_dir = system.file('extdata', 'rna_basecall_1D_001', package = 'tailfindr'),
                 save_dir = '~/Downloads',
                 csv_filename = 'rna_tails.csv',
                 num_cores = 2,
                 basecall_group = 'Basecall_1D_001',
                 save_plots = TRUE,
                 plot_debug_traces = TRUE,
                 plotting_library = 'rbokeh')
# N.B.: tailfindr cannot work if it can't find Events or Move table in
# your FAST5 files. MinKNOW Live Basecalling currently does not save the
# Events/Move table in the FAST5 file. If your reads have been live
# basecalled, then you should rebasecall them using Albacore or Guppy, and
# subsequently use tailfindr and specify the basecall_group parameter. Most
# probably, in the second round of your basecalling, the Events/Move table
# is stored in the 'Basecall_1D_001' section, so set this as the value of the
# basecall_group parameter. You can also confirm this by viewing your
# re-basecalled reads in HDFView.

## End(Not run)

adnaniazi/tailfindr documentation built on March 23, 2024, 7:07 a.m.