read.dna: Read DNA Sequences in a File

Description

These functions read DNA sequences in a file, and returns a matrix or a list of DNA sequences with the names of the taxa read in the file as rownames or names, respectively. By default, the sequences are returned in binary format, otherwise (if as.character = TRUE) in lowercase.

Usage

read.dna(file, format = "interleaved", skip = 0,
         nlines = 0, comment.char = "#",
         as.character = FALSE, as.matrix = NULL)
read.FASTA(file, type = "DNA")
read.fastq(file, offset = -33)

Value

a matrix or a list (if format = "fasta") of DNA sequences stored in binary format, or of mode character (if as.character = "TRUE").

read.FASTA always returns a list of class "DNAbin" or

"AAbin".

read.fastq returns a list of class "DNAbin" with an atrribute "QUAL" (see examples).

Arguments

file: a file name specified by either a variable of mode character, or a double-quoted string. Can also be a connection (which will be opened for reading if necessary, and if so closed (and hence destroyed) at the end of the function call). Files compressed with GZIP can be read (the name must end with .gz), as well as remote files.
format: a character string specifying the format of the DNA sequences. Four choices are possible: "interleaved", "sequential", "clustal", or "fasta", or any unambiguous abbreviation of these.
skip: the number of lines of the input file to skip before beginning to read data (ignored for FASTA files; see below).
nlines: the number of lines to be read (by default the file is read untill its end; ignored for FASTA files)).
comment.char: a single character, the remaining of the line after this character is ignored (ignored for FASTA files).
as.character: a logical controlling whether to return the sequences as an object of class "DNAbin" (the default).
as.matrix: (used if format = "fasta") one of the three followings: (i) NULL: returns the sequences in a matrix if they are of the same length, otherwise in a list; (ii) TRUE: returns the sequences in a matrix, or stops with an error if they are of different lengths; (iii) FALSE: always returns the sequences in a list.
type: a character string giving the type of the sequences: one of "DNA" or "AA" (case-independent, can be abbreviated).
offset: the value to be added to the quality scores (the default applies to the Sanger format and should work for most recent FASTQ files).

Author

Emmanuel Paradis and RJ Ewing

Details

read.dna follows the interleaved and sequential formats defined in PHYLIP (Felsenstein, 1993) but with the original feature than there is no restriction on the lengths of the taxa names. For these two formats, the first line of the file must contain the dimensions of the data (the numbers of taxa and the numbers of nucleotides); the sequences are considered as aligned and thus must be of the same lengths for all taxa. For the FASTA and FASTQ formats, the conventions defined in the references are followed; the sequences are taken as non-aligned. For all formats, the nucleotides can be arranged in any way with blanks and line-breaks inside (with the restriction that the first ten nucleotides must be contiguous for the interleaved and sequential formats, see below). The names of the sequences are read in the file. Particularities for each format are detailed below.

Interleaved: the function starts to read the sequences after it finds one or more spaces (or tabulations). All characters before the sequences are taken as the taxa names after removing the leading and trailing spaces (so spaces in taxa names are not allowed). It is assumed that the taxa names are not repeated in the subsequent blocks of nucleotides.
Sequential: the same criterion than for the interleaved format is used to start reading the sequences and the taxa names; the sequences are then read until the number of nucleotides specified in the first line of the file is reached. This is repeated for each taxa.
Clustal: this is the format output by the Clustal programs (.aln). It is close to the interleaved format: the differences are that the dimensions of the data are not indicated in the file, and the names of the sequences are repeated in each block.
FASTA: this looks like the sequential format but the taxa names (or a description of the sequence) are on separate lines beginning with a `greater than' character `>' (there may be leading spaces before this character). These lines are taken as taxa names after removing the `>' and the possible leading and trailing spaces. All the data in the file before the first sequence are ignored.

The FASTQ format is explained in the references.

Compressed files must be read through connections (see examples). read.fastq can read compressed files directly (see examples).

References

Anonymous. FASTA format. https://en.wikipedia.org/wiki/FASTA_format

Anonymous. FASTQ format. https://en.wikipedia.org/wiki/FASTQ_format

Felsenstein, J. (1993) Phylip (Phylogeny Inference Package) version 3.5c. Department of Genetics, University of Washington. http://evolution.genetics.washington.edu/phylip/phylip.html

Examples

Run this code

## 1. Simple text files

TEXTfile <- tempfile("exdna", fileext = ".txt")

## 1a. Extract from data(woodmouse) in sequential format:
cat("3 40",
"No305     NTTCGAAAAACACACCCACTACTAAAANTTATCAGTCACT",
"No304     ATTCGAAAAACACACCCACTACTAAAAATTATCAACCACT",
"No306     ATTCGAAAAACACACCCACTACTAAAAATTATCAATCACT",
file = TEXTfile, sep = "\n")
ex.dna <- read.dna(TEXTfile, format = "sequential")
str(ex.dna)
ex.dna

## 1b. The same data in interleaved format, ...
cat("3 40",
"No305     NTTCGAAAAA CACACCCACT",
"No304     ATTCGAAAAA CACACCCACT",
"No306     ATTCGAAAAA CACACCCACT",
"          ACTAAAANTT ATCAGTCACT",
"          ACTAAAAATT ATCAACCACT",
"          ACTAAAAATT ATCAATCACT",
file = TEXTfile, sep = "\n")
ex.dna2 <- read.dna(TEXTfile)

## 1c. ... in clustal format, ...
cat("CLUSTAL (ape) multiple sequence alignment", "",
"No305     NTTCGAAAAACACACCCACTACTAAAANTTATCAGTCACT",
"No304     ATTCGAAAAACACACCCACTACTAAAAATTATCAACCACT",
"No306     ATTCGAAAAACACACCCACTACTAAAAATTATCAATCACT",
"           ************************** ******  ****",
file = TEXTfile, sep = "\n")
ex.dna3 <- read.dna(TEXTfile, format = "clustal")

## 1d. ... and in FASTA format
FASTAfile <- tempfile("exdna", fileext = ".fas")
cat(">No305",
"NTTCGAAAAACACACCCACTACTAAAANTTATCAGTCACT",
">No304",
"ATTCGAAAAACACACCCACTACTAAAAATTATCAACCACT",
">No306",
"ATTCGAAAAACACACCCACTACTAAAAATTATCAATCACT",
file = FASTAfile, sep = "\n")
ex.dna4 <- read.dna(FASTAfile, format = "fasta")

## The 4 data objects are the same:
identical(ex.dna, ex.dna2)
identical(ex.dna, ex.dna3)
identical(ex.dna, ex.dna4)

## 2. How to read GZ compressed files

## create a GZ file and open a connection:
GZfile <- tempfile("exdna", fileext = ".fas.gz")
con <- gzfile(GZfile, "wt")
## write the data using the connection:
cat(">No305", "NTTCGAAAAACACACCCACTACTAAAANTTATCAGTCACT",
    ">No304", "ATTCGAAAAACACACCCACTACTAAAAATTATCAACCACT",
    ">No306", "ATTCGAAAAACACACCCACTACTAAAAATTATCAATCACT",
    file = con, sep = "\n")
close(con) # close the connection

## read the GZ'ed file:
ex.dna5 <- read.dna(gzfile(GZfile), "fasta")

## This example is with a FASTA file but this works as well
## with the other formats described above.

## All 5 data objects are identical:
identical(ex.dna, ex.dna5)

unlink(c(TEXTfile, FASTAfile, GZfile)) # clean-up

if (FALSE) {
## 3. How to read files from a ZIP archive

## NOTE: since ape 5.7-1, all files in these examples are written
## in the temporary directory, thus the following commands work
## best when run in the user's working directory.

## write the woodmouse data in a FASTA file:
data(woodmouse)
write.dna(woodmouse, "woodmouse.fas", "fasta")
## archive a FASTA file in a ZIP file:
zip("myarchive.zip", "woodmouse.fas")
## Note: the file myarchive.zip is created if necessary

## Read the FASTA file from the ZIP archive without extraction:
wood2 <- read.dna(unz("myarchive.zip", "woodmouse.fas"), "fasta")

## Alternatively, unzip the archive:
fl <- unzip("myarchive.zip")
## the previous command eventually creates locally
## the fullpath archived with 'woodmouse.fas'
wood3 <- read.dna(fl, "fasta")

identical(woodmouse, wood2)
identical(woodmouse, wood3)
}

## read a FASTQ file from 1000 Genomes:
if (FALSE) {
a <- "https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/phase3/data/HG00096/sequence_read/"
file <- "SRR062641.filt.fastq.gz"
URL <- paste0(a, file)
download.file(URL, file)
## If the above command doesn't work, you may copy/paste URL in
## a Web browser instead.
X <- read.fastq(file)
X # 109,811 sequences
## get the qualities of the first sequence:
(qual1 <- attr(X, "QUAL")[[1]])
## the corresponding probabilities:
10^(-qual1/10)
## get the mean quality for each sequence:
mean.qual <- sapply(attr(X, "Q"), mean)
## can do the same for var, sd, ...
}

Run the code above in your browser using DataLab