Korilog and Inria announce collaboration on significant acceleration of BLAST

Questembert, France, June 15th, 2012 - Korilog and Inria today announced the integration of the PLAST technology into Korilog software platforms. Licensed by Korilog from Inria, PLAST is a fast, accurate and NGS scalable bank-to-bank sequence similarity search tool providing significant accelerations of the Blast suite of algorithms.

Korilog and the Inria Genscale’s team collaborated closely on a eighteen-month R&D project to create KLAST, a new optimized implementation of the PLAST algorithm recently published in BMC Bioinformatics.

"A key issue in searching for similarities between sequence databanks is dramatically reducing execution time, while keeping a high quality in the resulting data. KLAST not only meet these two goals but its innovative implementation takes into account the exisiting IT infrastructure of many labs. Indeed, KLAST achieves accelerations without the need of any additional hardware devices, but it takes full advantage of multi-core processors available in common desktop computers or cluster nodes. We are very excited to provide our users with such an elegant tool", states Patrick Durand, CEO at Korilog.

KLASTp benchmarks using the black cottonwood Populus trichocarpa species have demonstrated a 24x increase in speed using KLAST software accelerator compared to BLAST, both tools ran on 8 cores of an Apple MacPro computer. This benchmark used KLASTp to compare 2,327 proteins against 2.9 million proteins from the NCBI RefSeq databank. Compared to BLAST and SSearch, KLAST is as sensitive and selective as these reference tools. Equivalent accelerations and quality are achieved with the other KLAST comparison tools, KLASTx, tKLASTn, tKLASTx and KLASTn.

Dominique Lavenier, PLAST scientific leader, notes: "The transfer of PLAST into a Hi-Tech company like Korilog is a very gratifying and important step for our research team. It allows us to directly contribute to new advances in emerging biological fields, from health and disease to agriculture and environment. And feedback from end-users is highly valuable to keeping PLAST as a technology leader in intensive sequence comparison." .

The KLAST sequence similarity search engine provides the following features:

• support for all types of sequence comparisons
• handle all lengths of sequences
• direct support of sequence files, without the need to pre-built static sequence indexes
• integrate a data filtering engine capable of selecting relevant hits with user-defined criteria (E-Value, identity, coverage, alignment length, etc.)
• NGS scalable

To provide users with an advanced sequence similarity search platform, the KLAST engine has been integrated in the Korilog Bioinformatics Extensions for KNIME. As a result, users benefit from Korilog plugin’s features and from a wide rande of data analysis methods coming with the KNIME platform. KLAST is available for the various releases of KNIME, from Desktop up to Cluster Execution. Command-line execution is also available.

"The integration of KLAST within the KNIME platform provides researchers with a very versatile genomics data analysis platform combining KNIME's extended data analysis nodes with the fast sequence similarity search engine created by Inria and Korilog. Using KLAST within KNIME will enable convenient sequence similarity results analysis, such as statistics, data filtering, data comparison, biological classification overview and interactive views", comments Michael Berthold, CEO at KNIME.AG.

Other enhancements are underway to speedup KLAST even more.

The KLAST project is part of the KoriPlast Collaborative Research program conducted by Korilog, financed by the Brittany Region and accompanied by CRITT Santé Bretagne.

During Summer and Fall 2012, Korilog will release new versions of its Bioinformatics Plugin for KNIME, KoriBlast and KoriBlast Server integrating the new high-performance KLAST search tool.

Stay in touch by following us on Twitter.

About Korilog

Korilog is a bioinformatics company providing next generation of graphical softwares to integrate and explore pathways, genomes and proteomes data. The company provides academic and industry research labs with desktop and web-based softwares (KoriBlast, KoriViewer) as well as expert software development services to support researchers in their R&D activities.

About Inria

Working at the crossroads of computer sciences and mathematics, over the last 40 years Inria’s researchers have been developing the scientific foundations for a new field of learning: computational sciences. When associated with other scientific disciplines, computational sciences can be used to offer new concepts, languages, methods and teaching aids which open up new avenues for exploration and understanding of complex phenomena. Working in project-teams, Inria researchers mix fundamental and applied research in an innovative blend to produce their results. The institute’s 174 teams, the majority of which are joint teams with other major French or international research bodies, are comprised of around twenty researchers working on a shared project for a period of 4 to 8 years. Inria is France’s only public research body fully dedicated to computational sciences. It hosts over 1000 young researchers each year.