Scroll to navigation

CENTRIFUGE(1) User Commands CENTRIFUGE(1)

NAME

centrifuge - rapid and memory-efficient system for classification of DNA sequences

DESCRIPTION

Centrifuge version 1.0.3-beta by the Centrifuge developer team (centrifuge.metagenomics@gmail.com) Usage:

centrifuge [options]* -x <cf-idx> {-1 <m1> -2 <m2> | -U <r>} [-S <filename>] [--report-file <report>]
<cf-idx>
Index filename prefix (minus trailing .X.cf).
<m1>
Files with #1 mates, paired with files in <m2>. Could be gzip'ed (extension: .gz) or bzip2'ed (extension: .bz2).
<m2>
Files with #2 mates, paired with files in <m1>. Could be gzip'ed (extension: .gz) or bzip2'ed (extension: .bz2).
<r>
Files with unpaired reads. Could be gzip'ed (extension: .gz) or bzip2'ed (extension: .bz2).
<filename>
File for classification output (default: stdout)
<report>
File for tabular report output (default: centrifuge_report.tsv)
<m1>, <m2>, <r> can be comma-separated lists (no whitespace) and can be specified many times. E.g. '-U file1.fq,file2.fq -U file3.fq'.

Options (defaults in parentheses):

Input:
query input files are FASTQ .fq/.fastq (default)
query input files are in Illumina's qseq format
query input files are (multi-)FASTA .fa/.mfa
query input files are raw one-sequence-per-line
<m1>, <m2>, <r> are sequences themselves, not files
skip the first <int> reads/pairs in the input (none)
stop after first <int> reads/pairs (no limit)
-5/--trim5 <int>
trim <int> bases from 5'/left end of reads (0)
-3/--trim3 <int>
trim <int> bases from 3'/right end of reads (0)
qualities are Phred+33 (default)
qualities are Phred+64
qualities encoded as space-delimited integers
treat all quality values as 30 on Phred scale (off)
do not align forward (original) version of read (off)
do not align reverse-complement version of read (off)

Classification:

minimum length of partial hits (default 22, must be greater than 15)
minimum summed length of partial hits per read (default 0)

--host-taxids <taxids> comma-separated list of taxonomic IDs that will be preferred in classification

--exclude-taxids <taxids> comma-separated list of taxonomic IDs that will be excluded in classification

Output:
define output format, either 'tab' or 'sam' (tab)
columns in tabular format, comma separated default: readID,seqID,taxID,score,2ndBestScore,hitLength,queryLength,numMatches
print wall-clock time taken by search phases
write unpaired reads that didn't align to <path>
write unpaired reads that aligned at least once to <path>
write pairs that didn't align concordantly to <path>
write pairs that aligned concordantly at least once to <path>
(Note: for --un, --al, --un-conc, or --al-conc, add '-gz' to the option name, e.g. --un-gz <path>, to gzip compress output, or add '-bz2' to bzip2 compress output.) --quiet print nothing to stderr except serious errors --met-file <path> send metrics to file at <path> (off) --met-stderr send metrics to stderr (off) --met <int> report internal counters & metrics every <int> secs (1)
Performance:

-o/--offrate <int> override offrate of index; must be >= index's offrate

-p/--threads <int> number of alignment threads to launch (1)

use memory-mapped I/O for index; many instances can share
Other:
filter out reads that are bad according to QSEQ filter
seed for random number generator (0)

--non-deterministic seed rand. gen. arbitrarily instead of using read attributes

print version information and quit
print this usage message

AUTHOR

This manpage was written by Andreas Tille for the Debian distribution and can be used for any other usage of the program.

October 2021 centrifuge 1.0.3