LASER: Large genome ASsembly EvaluatoR

email

Download

 

 

Overview:

LASER is a new program for large genome assembly evaluation.

LASER is based on QUAST genome assembly program and provides huge performance improvements for large genomes. LASER is 5.6 times faster and requires only half of the memory required by QUAST program for Human genome assemblies. See comparisons.

The improvements in LASER are obtained by primarily replacing MUMmer MEMs (Maximal Exact Matches) computation program with E-MEM (Efficient computation of maximal exact matches for large genomes). E-MEM is the best program for MEMs computation both in terms of memory and run time performance. E-MEM was published in Bioinformatics in 2014 and a copy of E-MEM program can be requested here.

Apart from using E-MEM, the NUCmer script (one of the scripts from MUMmer package) is also modified for improving performance. Lastly, some changes have been made to existing QUAST code to reduce the overall memory requirements. There is no change in functionality or command line options from its predecessor.

The current version of LASER is based on QUAST version 2.3 and MUMmer version 3.23. Similar to QUAST, the LASER utilizes other programs such as GeneMarkS, MetaGeneMark, GeneMark-ES, GlimmerHMM and GAGE. These programs have not been modified or altered in anyways in LASER code.

The current version of LASER was released under GNU GPL version 3 ( see LICENSE ).

Note: We named our program initially E-QUAST since it uses a large part of the QUAST source code but later changed the name at the request of QUAST's authors. They are preparing several other programs with "QUAST" in the name and another one would have created confusion.

Benefits of LASER over QUAST:

The biggest benefit of LASER is its runtime performance. Our test results on Human genome assemblies show that LASER program is typically 5.6 times faster and requires only half the memory required by QUAST.

comparison
The assemblies used for comparison, described in the paper, are available for download here: H1, H2, H3, H4, H5.

 

Installation:

LASER program is customized for Linux and Mac OS. The program requires g++, make, python (2.6 higher), perl and openMP.

The installation starts by downloading file LASER.zip.

Unzip file using command "unzip LASER.zip". This will create a directory LASER and all required files in this folder.

As suggested on QUAST manual page, please run the quast/metaquast python scripts in test mode to check the correctness of the installation.

This can be done using following commands:

python quast.py --test 
python metaquast.py --test  

The root LASER installation folder contains a file "clean". This is a shell script which can be used to clean your installation. The script will clean all your dependencies and also compiles all the required binaries. The script is also useful in case an existing installation is moved, copied or renamed. In such cases, simply run "clean" file to make installation up-to-date and ready to use.

 

Running LASER:

LASER can be run with exactly same command line options as QUAST. The detailed description of all the command line options can be found at QUAST manual page.

The example included with LASER installation can be run as:

./quast.py test_data/contigs_1.fasta test_data/contigs_2.fasta \
           -R test_data/reference.fasta.gz -G test_data/genes.txt \
           -O test_data/operons.txt


The output of LASER program can be viewed in file:  ./quast_results/latest/report.txt

 

Authors:

Nilesh Khiste, Lucian Ilie (ilie@uwo.ca)
Department of Computer Science
University of Western Ontario
London, ON N6A 5B7
Canada

web page hit counters codes Free
Site Hits Total