By Matt Fates, Partner and Jeff Knox, Associate
The last two decades have seen tremendous expansion of biological datasets and our ability to generate them, enabled among other things by advances in Next-Generation Sequencing (NGS). As we generate more information faster and cheaper than ever before, the ability to study that data via bioinformatics – efficiently and reproducibly – becomes paramount. While the role of bioinformaticians will only grow, we’ll increasingly need tools that allow biologists to perform independent analysis of their own complex data.
We at Innospark believe no one’s built a solution that does the job like Via Scientific. Via’s Foundry platform is now used by companies small and large. It includes AI-based features that empower biologists to perform and repeat bioinformatic analysis, better manage metadata and compute, operate across a range of analysis types. With their potential to accelerate our increasing store of biology data toward new discoveries, Via is well aligned with Innospark’s drive to use AI for good.
The Promise and Challenge of Expanding Biology Data
The advent of Next-Generation Sequencing (NGS) has seen breakthroughs for our ability to generate biological data in quantity. The massively parallel methods offered by Illumina and its peers have enabled new types of analysis in several fields – whole genome sequencing in genomics, RNA sequencing in transcriptomics, epigenomic analyses, the list goes on – and the pace of growth is astounding. Genomics data alone is doubling every seven months. The wealth of new biological data promises to enable disease research and the development of new therapies and other biological products, but only if we can analyze it effectively.
The answer of course is bioinformatics, the application of computation to collect and analyze biological data. Bioinformatics is not a new discipline. Some trace the field back to the mid-60s, when the results of early protein sequencing experiments were organized into databases (shoutout to you, Margaret Dayhoff). Just as the field evolved in response to new technology then, the same is happening again in response to NGS technologies. Beginning in the mid-2000’s bioinformatics graduate programs sprang up at major universities. Bioinformatics companies have been built to help researchers manage their reems of data. What was a discipline is fast becoming an industry.
Even so, with the substantial acceleration of data growth, efficient analysis remains a major challenge for both biopharma and academia. We still don’t have enough bioinformaticians. Biologists are highly trained, but not as data or computer scientists. As biologists increasingly need major compute resources and seek out data science professionals to help them analyze data, a real bottleneck has formed. This bioinformatics bottleneck limits the speed at which data intensive biological research can be performed.
The Via Solution
We’re excited about Via Scientific because we believe they’re the platform that will let scientists go faster. Perhaps the best solution to the bioinformatics bottleneck is to give self-sufficiency back to the biologist. Researchers must have the tools to reuse common bioinformatic analysis programs, to make modifications without the help of the original author, and to perform routine analysis on their own.
Foundry, Via’s bioinformatics platform solution, was originally developed within the BioInformatics Core at UMass Chan Medical School. CTO Alper Kucukural developed the first versions with Drs. Melissa Moore (then professor, later Moderna CSO) and Manuel Garber (Chan, Broad) to meet their own research needs, and eventually the multi-faceted research needs of the entire UMass Chan research community. Foundry allows bioinformatics analysis algorithms, referred to as pipelines, to be written by bioinformaticians and then subsequently modified by biologists without intense coding. Foundry also comes loaded with a sizeable library of standard pipelines. With the platform’s help, biologists can perform routine bioinformatics analysis independently. Bioinformaticians are freed to construct novel pipelines or perform especially complex analysis.
Importantly, the Foundry platform builds in solutions to problems that might arise from the decentralization it creates. Some will use AI in ways we at Innospark find exciting. Foundry captures and manages metadata well, it seamlessly sets up computational environments, it includes clear management of versioning – if consistency of analysis is the silver lining of a limited number of bioinformaticians, Foundry tries to get ahead of the issues that might crop up when you make the number of people performing analysis a lot bigger.
By bringing bioinformatics analysis back to the biologist and accelerating the speed of data-intensive science, Via Scientific has already proven its value. We’re excited by the growing list of pharma, biotech, and academic customers taking advantage of Foundry.
Into the future of distributed bioinformatics
Via’s Foundry was first built to help a few research labs to perform their bioinformatics analysis faster. It’s grown to meet the needs of a center, some friendly organizations, and now – after years of iterative product testing within academia and with industry partners – it’s ready to serve everyone. We believe it will enable scientists to better take advantage of the wealth of data generated in wet labs today and move us more quickly toward the lifesaving treatments and other biology advances that data will inspire tomorrow.
Innospark is excited to team up with G20 Ventures to back the Via team, led by the experienced hand of its CEO Jim Crowley and the technical expertise of its first architect CTO Alper Kucukural, and honored to join Melissa Moore in helping to advise the company on the board.
Bioinformatics education: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1289384/