System Downloads
System Requirements
- Java 2 Runtime Environment, Standard Edition v1.4 (or greater)
- NOTE:We have received reports of problems running the distributed system on some non Sun JVM's - therefore we recommend you run the system on a Sun JVM if possible. To check your JVM version, type java -version on the command line.
Distributed System
- v2.0.2 November 2005
- fixed bug with dynamic memory requirement of units matching suitable client machines. - v2.0.1 October 2005
- Full multi-user accounts added
- Implemented usergroups with group job priorities
- Administrator can send messages to users via the system
- Fixed many minor bugs in server and client
- NOTE: Result zip files fail to open with earlier versions of Winzip (pre 7.0). We recommend Windows users to use 7-zip to open result files. - v1.4 May 2005
- Individual units can now dynamically adapt during a computation in terms of memory and CPU requirements
- Remote interface now displays more detailed information about result files for easier selection of results
- Result files are automatically deleted after 10 days
- Users now enter their name when entering a job
- Full multi user support coming in next update! - v1.3 Jan 2005
- reduced memory usage of server, fixed a bug in the client, improved usability of the remote interface - v1.2 Jun 2004
- fixed a number of bugs and updated to changes in programming api (upgrade recommended) - v1.1 Mar 2004
- fixed a few minor bugs on client and server (upgrade recommended) - v1.0 Jan 2004
- first release of the distributed system (upgrade recommended)
Multi-Tiered Distributed System
- v1.0beta Jan 2004
- first beta release of the multi-tiered version of the distributed system
Application Downloads
MultiPhyl
- v1.0.6 Jul 2006
- fixed problem where too many threads (for creating distance matrices) were being created on the server. - v1.0.5 Feb 2006
- fixed minor bug when reading phylip alignments with very short names
- speedup in NJ tree construction for likelihood tree searches - v1.0.4 Dec 2005
- added ML base frequency optimisation to all nucleotide models
- we were notified of a bug in the PAL library limiting branch lengths to a maximum length of 1.0 (now increased to 100.0)
- fixed bug causing a few very short amino acid alignments to be read as nucleotides - v1.0.3 22nd Nov 2005
- fixed bug causing tree not to be printed to file when user chooses to use a particular nucleotide model for their analyses
- source code is distributed with the application (source folder) - v1.0.2 4th Nov 2005
- fixed bug causing model selection to always choose the BIC criterion in certain alignments - v1.0.1 19th Oct 2005
- optimised code of v1.0 to reduce memory usage on the server
- fixed bug causing a few small alignments to not converge properly - v1.0 1st Oct 2005
- first release of MultiPhyl - Keane, T.M., Naughton, T.J., Travers, S.A.A., McInerney, J.O., McCormack, G.P. (2005) DPRml: Distributed Phylogeny Reconstruction by Maximum Likelihood, Bioinformatics, 21(7):969-974 - more info
- v1.0 Jul 2004
- first release of DSEARCH - Keane, T.M. and Naughton, T.J. (2005) DSEARCH: sensitive database searching using distributed computing, Bioinformatics, 21(8):1705-1706 - more info
- v1.1 Stable Jun 2004
- name of program changed to DPRml
- fixed problem in tree rearrangements that caused a small proportion of tree not to be issued - v1.0LTC Apr 2004
- added compression that enables the program to build much larger trees - v1.0 Jan 2004
- first release of DCPml - Keane, T.M., Naughton, T.J., Travers, S.A.A., McInerney, J.O., McCormack, G.P. (2005) DPRml: Distributed Phylogeny Reconstruction by Maximum Likelihood, Bioinformatics, 21(7):969-974 - more info
If you use MultiPhyl please cite:
DSEARCH
NOTE: The current version of DSEARCH will only work with versions 1.X of the distributed system. An update will be posted soon to make it compliant with version 2.0 of the system.If you use DSEARCH please cite:
DPRml
If you use DPRml please cite:
Sample problems
We have produced a number of sample distributed computations (sorting, substring matching, character occurrence counting etc...) for testing purposes which are available in a self-contained zip file. These are ideal examples for learning how to create problems to run on the distributed system.
Recent Publications
Distributed Monte Carlo Simulation of Light Transportation in TissueApril 2006
This paper is to appear in at 8th International Workshop on Java for Parallel and Distributed Computing.
Framework for task scheduling in heterogeneous distributed computing using genetic algorithmsNovember 2005
This journal paper is to appear in Artificial Intelligence Review. It describes a distributed task scheduling scheme based on genetic algorithms.
DPRml: Distributed Phylogeny Reconstruction by Maximum LikelihoodMarch 2005
This journal paper appeared in Bioinformatics. It describes a distributed phylogeny application.
DSEARCH: sensitive database searching using distributed computingMarch 2005
A distributed bioinformatics database searching application that uses the most accurate search algorithms. It is to appear in Bioinformatics.
Adaptive Scheduling Across a Distributed Computation PlatformJune 2004
This paper was presented at ISPDC'04, and describes an adaptive scheduler for a Java distributed computation system.