import_reanno: Imports annotation from reannotation mapping

In Danis102/seqpac: Seqpac: A Framework for smallRNA analysis in R using Sequence-Based Counts

Imports annotation from reannotation mapping

Description

This function imports bowtie output and summarizes the content.

Usage

import_reanno(
  bowtie_path,
  threads = 1,
  coord = FALSE,
  report = "minimum",
  reduce = NULL
)

Arguments

bowtie_path	Path to a directory where bowtie output files can be found.
threads	Integer stating the number of parallel jobs.
coord	Logical whether or not mapping coordinates should be reported when report="full".
report	Character vector indicating what to report "minimum" or "full" (default="minimum").
reduce	Character indicating a reference name (without file type extension) that should be exempted from report="full" and instead will generate a minimum report.

Details

Given the path to alignment outputs from bowtie, import_reanno will attempt to read these files into R and generate a list of unsorted data.frames where each row summarizes all annotations for a given sequence. It is called by map_reanno to generate summarized data from bowtie output saved as .Rdata files. These files can then be re-imported into R where information is organized using make_reanno and finally added to the annotation table (PAC$Anno) of the original PAC-list object using add_reanno.

Value

List of data frames with additional information from reannotation files generated with bowtie. If report="minimum", the function will report each hit only as the number of mismatches for a given reference file. If report="full" the full name reported in the fasta file used as reference in the bowtie reannotation will be reported. If a reference name is specified in reduce, this reference is excerpted from report="full" and is instead reported as report="minimum".

Caution: Large references with lots of redundancy (such as pirBase in some species) will cause massive character strings if report="full" with no restrictions. Specifying such references in reduce=<reference_names> will circumvent this problem.

Important

Re-annotation must have been done using Bowtie default output settings, where output files specifically have been named '.txt'. The function will not work with other formats such as SAM/BAM formats.

The basenames of the bowtie output will also be used to annotate. Example: bowtie -v 3 -a -f '<piRBase.fasta>' '<master_anno.fasta>' piRNA.txt Will annotate sequences as piRNA since the txt-file output was named 'piRNA'

See Also

http://bowtie-bio.sourceforge.net/index.shtml for information about Bowtie and for Rbowtie: https://www.bioconductor.org/packages/release/bioc/html/Rbowtie.html. https://github.com/Danis102 for updates on the current package.

Other PAC reannotation: add_reanno(), as.reanno(), make_conv(), make_reanno(), map_reanno(), simplify_reanno()

Examples

######################################################### 
##### Test import_reanno

### First, if you haven't already generated Bowtie indexes for the included
# fasta reference you need to do so. If you are having problem see the small
# RNA guide (vignette) for more info.

## tRNA reference:
trna_file <- system.file("extdata/trna", "tRNA.fa", 
                         package = "seqpac", mustWork = TRUE)
trna_dir<- gsub("tRNA.fa", "", trna_file)

if(!sum(stringr::str_count(list.files(trna_dir), ".ebwt")) ==6){
  Rbowtie::bowtie_build(trna_file, 
                        outdir=trna_dir, 
                        prefix="tRNA", force=TRUE)
}
##  Then load a PAC-object and remove previous mapping from anno:
load(system.file("extdata", "drosophila_sRNA_pac_filt_anno.Rdata", 
                 package = "seqpac", mustWork = TRUE))

ref_paths <- list(trna= trna_file)

##  You may add an output path of your choice, but here we use a temp folder:
output <- paste0(tempdir(),"/seqpac/test")

##  Then map the PAC-object against the fasta references. Warning: if you use
# your own data, you may want to use override=FALSE, to avoid deleting previous
# mapping by mistake. keep_temp=TRUE can be used to run import_reanno
# independently.

map_reanno(pac, ref_paths=ref_paths, output_path=output,
           type="internal", mismatches=2,  import="biotype", 
           threads=2, keep_temp=TRUE, override=TRUE)

reanno1 <- import_reanno(output, report="Biotype",  threads=1)

Danis102/seqpac documentation built on Aug. 26, 2023, 10:15 a.m.