Biocceleration Releases GenCore 6.2,
with Support for Percent Identity Searches and MPI based Parallelism
March 16, 2007 – Biocceleration Ltd. announced the release of GenCore
6.2, an updated version of its advanced sequence search acceleration
system.
GenCore 6.2 offers the following new features and capabilities:
- A new Percent Identity Search
algorithm, that is best suited for
intellectual property
analysis. This algorithm calculates the best percent
identity between the query sequence and the database sequence, and
reports the database sequences that achieved the highest percent
identity score against the query.
- Using MPI based
implementation, GenCore dynamic programming
search applications
can now be parallelized over a cluster of Linux nodes. This enables
running a single database search over a large number of Linux
servers, thus providing orders of magnitude acceleration for
computationally intensive searches.
These new capabilities are in addition to the already rich feature set
of GenCore 6:
- A robust software package that runs on standard, general
purpose computing and network architectures (Intel Pentium-4 or Opteron
processor
based systems or others), using Linux or other Unix operating systems.
-
Software based acceleration, using Biocceleration's highly optimized
proprietary software implementation.
-
Up to 10-fold speed improvement relative to standard software
implementations. More then 2 Billion Smith-Waterman Matrix cells
calculations per second on a dual 3GHz P4 Linux machine. For detailed
benchmarks see:
http://www.biocceleration.com/GenCore6-Benchmarks.html
-
Support for a comprehensive set of search algorithms including
Smith-Waterman (with affine scoring), ProfileSearch, ProfileScan,
Translated Searches, Frame+, ProfileFrame+, HMMsearch, HMMscan, GeneWise
-
Seamless integration with a wide variety of DNA and protein database
formats including FASTA, BLAST, GenBank, EMBL, SwissProt, GCG, NBRF,
PIR, RAW, IG, PFAM (HMM)
-
Full user control over search parameters such as gap penalties,
comparison tables, translation tables, scoring and normalization
methods, and others.
-
Multi-query operation enables multiple sequences, or even a database
subset, to be submitted as the search query set.
-
Multi-threaded operation enables full utilization of multiprocessor
Unix/Linux machines.
-
Band, linear memory alignment algorithm for all search applications,
produces fast, accurate alignment, even for very long sequences.
-
Support for multiple normalization methods enables customizing the
software output for specific research projects.
With free download license for academic and research institutes, and
reasonably priced license for commercial entities, Biocceleration
enables existing Bioccelerator customers (BioXL/G, BioXL/P, BioXL/H),
and new customers, easy upgrade to a new level of database search
speed, performance, and flexibility.
For additional information,
please visit Biocceleration’s products page at: http://www.biocceleration.com/Products.html
or contact us at info@biocceleration.com
Biocceleration
Ltd. , a
bioinformatics company, is the leading provider of systems and
solutions for accelerating searches in protein and nucleic acids
databases. Biocceleration provides software and hardware solutions that
accelerate rigorous database search algorithms, like HMMsearch and
Smith-Waterman by a factor of 10-1000 relative to general purpose
servers. With install
base of more then 50 biotech companies, pharmaceutical companies, and
academic research institutes, Biocceleration offers a suite of hardware
accelerators and software packages to support the industry needs for
accelerated protein and DNA database searches.
.
|