Recent advances in next-generation sequencing technologies require alignment algorithms and software that can keep pace with the heightened data production. Standard algorithms, especially protein similarity searches, represent significant bottlenecks in analysis pipelines. For metagenomic approaches in particular, it is now often necessary to search hundreds of millions of sequence reads against large databases. Here we describe mBLAST, an accelerated search algorithm for translated and/or protein alignments to large datasets based on the Basic Local Alignment Search Tool (BLAST) and retaining the high sensitivity of BLAST. The mBLAST algorithms achieve substantial speed up over the National Center for Biotechnology Information (NCBI) programs BLASTX, TBLASTX and BLASTP for large datasets, allowing analysis within reasonable timeframes on standard computer architectures. In this article, the impact of mBLAST is demonstrated with sequences originating from the microbiota of healthy humans from the Human Microbiome Project. mBLAST is designed as a plug-in replacement for BLAST for any study that involves short-read sequences and includes high-throughput analysis. The mBLAST software is freely available to academic users at www.multicorewareinc.com.
This paper discusses the multiple-instruction stream/ multiple-data stream (MIMD) Wisconsin parallel array computer (WISPAC). The design of the computer makes use of current, low-cost microprocessor technology and is intended to meet future needs in large-scale simulation. A modular approach is taken where smaller configurations are meant as direct competitive replacements of current hybrid systems. Larger configurations will allow high-speed solution of large-scale simulation problems using a one-to-one correspondence of each microprocessor to each node of the system being modeled. A three-dimensional configuration is used to allow a high degree of isomorphism between the computer setup and the model to be investigated. These schemes simplify programming requirements and overall machine conceptualization. New techniques of high-speed internode communication utilizing a serial "pass-through" scheme are described. The scheme allows general cross-array communication that is competitive with cross-point methods. Initial work on a minimal parallel configuration is also described. The WISPAC authors, from left to right in the photo: Mike Redmond (BSEE, Wisconsin, 1976) is supervisor of the Data Acquisition and Simulation Laboratory (DASL) at Wisconsin. Currently, DASL construction facilities are being used in development of the WISPAC prototype.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.