Ray is a parallel de novo genome assembler that utilises the message-passing interface everywhere and is implemented using peer-to-peer communication.
Ray is implemented using RayPlatform, a message-passing-interface programming framework.
Ray is documented in
- Documentation/ (many files)
- MANUAL_PAGE.txt (command-line options, same as Ray -help)
- README.md (general)
- INSTALL.txt (quick installation)
- de novo genome assembly
- de novo meta-genome assembly (with Ray Méta)
- de novo transcriptome assembly (works, but not tested a lot)
- quantification of contig abundances
- quantification of microbiome consortia members (with Ray Communities)
- quantification of transcript expression
- taxonomy profiling of samples (with Ray Communities)
- gene ontology profiling of samples (with Ray Ontologies)
see Documentation/BiologicalAbundances.txt
If you want to contribute, clone the repository, make changes and I (Sébastien Boisvert) will pull from you after reviewing the code changes.
-
Users: denovoassembler-users AT lists.sourceforge.net
-
Read it on gmane: http://blog.gmane.org/gmane.science.biology.ray-genome-assembler
-
Development/hacking: denovoassembler-devel AT lists.sourceforge.net
-
SEQanswers: http://seqanswers.com/forums/showthread.php?t=4301
You need a C++ compiler (supporting C++ 1998), make, an implementation of MPI (supporting MPI 2.2).
tar xjf Ray-x.y.z.tar.bz2
cd Ray-x.y.z
make PREFIX=build
make install
ls build
tar xjf Ray-x.y.z.tar.bz2
cd Ray-x.y.z
mkdir build
cd build
cmake ..
make
make PREFIX=build2000 MPICXX=/software/openmpi-1.4.3/bin/mpicxx
make install
Tested C++ compilers: see Documentation/COMPILERS.txt
Some processors have the popcnt instruction and other cool instructions. With gcc, add -march=native to build Ray for the processor used for the compilation.
make PREFIX=Build.native DEBUG=n ASSERT=n EXTRA=" -march=native"
make install
The best way to build Ray is to use whole-program optimization. With gcc, use this script:
./scripts/Build-Link-Time-Optimization.sh
make PREFIX=Ray-Large-k-mers MAXKMERLENGTH=64
# wait
make install
mpirun -np 512 Ray-Large-k-mers/Ray -k 63 -p lib1_1.fastq lib1_2.fastq \
-p lib2_1.fastq lib2_2.fastq -o DeadlyBug,Assembler=Ray,K=63
# wait
ls DeadlyBug,Assembler=Ray,K=63/Scaffolds.fasta
make PREFIX=build-3000 MAXKMERLENGTH=64 HAVE_LIBZ=y HAVE_LIBBZ2=y \
ASSERT=n FORCE_PACKING=y
# wait
make install
ls build-3000
see the Makefile for more.
To run Ray on paired reads:
mpiexec -n 25 Ray -k31 -p lib1.left.fasta lib1.right.fasta -p lib2.left.fasta lib2.right.fasta -o RayOutput
ls RayOutput/Contigs.fasta
ls RayOutput/Scaffolds.fasta
ls RayOutput/
Ray can be run with a configuration file instead.
mpiexec -n 16 Ray Ray.conf
Content of Ray.conf:
-k 31 # this is a comment -p lib1.left.fasta lib1.right.fasta
-p lib2.left.fasta lib2.right.fasta
-o RayOutput
RayOutput/Contigs.fasta and RayOutput/Scaffolds.fasta
type Ray -help for a full list of options and outputs
Ray assembles color-space reads and generate color-space contigs. Files must have the .csfasta extension. Nucleotide reads can not be mixed with color-space reads. This is an experimental feature.
http://denovoassembler.sf.net/publications.html
cd code
doxygen DoxygenConfigurationFile
cd DoxygenDocumentation/html
firefox index.html
- http://aws.amazon.com/ec2/hpc-applications/
- https://cloud.genomics.cn/
- http://szdaily.sznews.com/html/2011-08/04/content_1689998.htm
- http://www.nature.com/nbt/journal/v28/n7/full/nbt0710-691.html
- http://dskernel.blogspot.com/2011/07/understand-main-loop-of-message-passing.html
- http://cw.squyres.com/
- http://blogs.cisco.com/performance/
- http://www.parawiki.org/index.php/Message_Passing_Interface#Peer_to_Peer_Communication
Doctoral Award to S.B., Canadian Institutes of Health Research (CIHR)
see AUTHORS
see Documentation/VISUAL_STUDIO.txt