CrocoBLAST:Motivation

From WebChem Wiki
Jump to: navigation, search

Everyone loves BLAST. We love BLAST too. What's not to love? BLAST is easy to understand and it still represents the gold standard in local alignment algorithms. But BLAST doesn't like parallel computing. Not even in the latest NCBI release. If your lab produces small data sets, you are probably familiar with the NCBI BLAST service. If you have medium or large data sets, perhaps you have access to a powerful computer and you can afford to wait a few weeks or even months to see whether your alignment is finished or you need to start over. Or perhaps you are among the lucky few who have access to a GPU machine equipped with a GPU BLAST code. Whatever your situation, if you are producing data from Next Generation Sequencing (NGS) experiments, you have likely encountered the limitations of currently available BLAST implementations. CrocoBLAST is for you.

CrocoBLAST offers a platform for planning, running, monitoring, and managing your BLAST calculations. With CrocoBLAST, you will always know how much time it will take to run a BLAST job, and you will be able to pause or interrupt a BLAST job at any time and retrieve partial results.

Another key aspect of CrocoBLAST is that it is extremely hungry. As such, it takes the sequences in the input file, and munches them into small pieces that will be fed into the classical BLAST algorithm. This enables you to run very large BLAST jobs efficiently even with minimal computational resources (say, your desktop machine), while ensuring that the output is identical to what you would obtain if you were to run BLAST (minus the headache and the constant frustration while you wait for the job to complete... or not).



Return to the Table of contents.