ntCard is a streaming algorithm for cardinality estimation in genomics datasets. As input it takes file(s) in fasta, fastq, sam, or bam formats and computes the total number of distinct k-mers, F0, and also the k-mer coverage frequency histogram, fi, i>=1.
Install Homebrew, and run the command
brew install brewsci/bio/ntcard
Install Linuxbrew, and run the command
brew install brewsci/bio/ntcard
When installing ntCard from GitHub source the following tools are required:
To generate the configure script and make files:
./autogen.sh
To compile and install ntCard in /usr/local:
$ ./configure
$ make
$ sudo make install
To install ntCard in a specified directory:
$ ./configure --prefix=/opt/ntCard
$ make
$ make install
ntCard uses OpenMP for parallelization, which requires a modern compiler such as GCC 4.2 or greater. If you have an older compiler, it is best to upgrade your compiler if possible. If you have multiple versions of GCC installed, you can specify a different compiler:
$ ./configure CC=gcc-xx CXX=g++-xx
For the best performance of ntCard, pass -O3
flag:
$ ./configure CFLAGS='-g -O3' CXXFLAGS='-g -O3'
To run ntCard, its executables should be found in your PATH. If you installed ntCard in /opt/ntCard, add /opt/ntCard/bin to your PATH:
$ PATH=/opt/ntCard/bin:$PATH
ntcard [OPTIONS] ... FILE(S) ...
Parameters:
-t
,--threads=N
: use N parallel threads [1] (N>=2 should be used when input files are >=2)-k
,--kmer=N
: the length of k-mer-c
,--cov=N
: the maximum coverage of k-mer in output[1000]
-p
,--pref=STRING
: the prefix for output file names-o
,--output=STRING
: the name for single output file name (can be used only for single compact output file)FILE(S)
: input file or set of files seperated by space, in fasta, fastq, sam, and bam formats. The files can also be in compressed (.gz
,.bz2
,.xz
) formats . A list of files containing file names in each row can be passed with@
prefix.
For example to run ntcard on a test file reads.fastq
with k=50
and output the histogram in a file with prefix freq
:
$ ntcard -k50 -p freq reads.fastq
To run ntcard on a test file reads.fastq
with multiple k's k=32,64,96,128
and output the histograms in files with prefix freq
use:
$ ntcard -k32,64,96,128 -p freq reads.fastq
As another example, to run ntcard on 5
input files file_1.fq.gz, file_2.fa, file_3.sam, file_4.bam, file_5.fq with k=64
and 6 threads and maximum frequency of c=100
on output file with prefix freq
:
$ ntcard -k64 -c100 -t6 -p freq file_1.fq.gz file_2.fa file_3.sam file_4.bam file_5.fq
If we have a list of input files lib.in
, to run ntCard with k=144
and 12
threads and output file with prefix freq
:
$ ntcard -k144 -t12 -p freq @lib.in
Output:
- The numbers
Fk
provide useful statistics on the input sequences. By default, F0 and F1 are output to stdout along with runtime.F0
denotes the number of distinct k-mers appearing in the stream sequencesF1
is the total number of k-mers in the input datasetsF2
is the Gini index of variation that can be used to show the diversity of k-mers andF∞
results in the most frequent k-mer in the input reads.
- A tab separated output file with columns
k
,f
, andn
.k
k-mer sizef
the frequency of a k-mern
the number of k-mers with frequencyf
Hamid Mohamadi, Hamza Khan, and Inanc Birol. ntCard: a streaming algorithm for cardinality estimation in genomics data. Bioinformatics (2017) 33 (9): 1324-1330. 10.1093/bioinformatics/btw832