October 15, 2013


How does DashO3 work?DashO3 logo

    • Provide us your analysis goals, and we provide you with a pipeline
    • Take an existing published analysis and create a custom pipeline
    • We identify computational and data flow bottlenecks in your existing pipelines and increase performance by 2-10x

Research-grade pipelines are not production-grade pipelines
Many bioinformatics pipelines that start as research projects eventually morph into production tools. Research projects, by their nature, tend to be quick and dirty implementations that get the job done. As they enter production, it is common to simply move research code to a production cluster and run it “as-is”. While this might initially work, as throughput increases, research pipelines can quickly become a computational bottleneck, limiting the turnaround time for processing samples. Of course, when this happens, the original developers might be working on new projects (or gone!) and can’t be called upon to speed things up.

We make your pipelines production-grade
The DashO3 (“dash oh-three”) program is designed to help when you need more performance from your pipelines. “-O3” is a standard compiler option that translates to: “use every technique at your disposal to make my program run fast.” We apply the same spirit to optimizing your pipelines for production. Using your existing pipelines, we leverage a collection of minimally invasive optimization methods to increase performance and meet your goals. Like using “-O3” on your individual programs, our process can often find optimizations that increase performance by 2-10x.


Download a DashO3 flyer and other literature.