Entrepreneurs found companies for many reasons. For Allen Chang, CEO of Atgenomix, family illnesses spurred him to pursue an approach to understanding and preventing disease. His company applies cloud technology and machine learning (ML) to create reproducible outcomes and deeper insights in genomics, thus accelerating the promise of precision medicine. The Tetra Partner Network is honored to feature him in our Spotlight Series of entrepreneurs.
Please tell us a little bit about your background.
I am an entrepreneur and technology innovator with over 18 years of experience in bioinformatics, cloud computing, and enterprise software products at a global scale. I have a degree in Computer Science from the University of Ottawa in Canada and have three cooperative peer-review publications.
Prior to founding Atgenomix, I was the senior product manager at Trend Micro, Inc. (a global leader in cloud and enterprise cybersecurity), managing three enterprise security product lines with over $30 million in revenue globally, including North and South America, EMEA, APAC, and Greater China. As the R&D manager for the Core Technology Group within Trend Micro, I led a team of talented software engineers and Ph.D. data scientists to develop cutting-edge, supervised, and unsupervised ML technologies in solving petabyte-scale big data problems and was granted seven US patents.
In a few sentences, please describe Atgenomix and why you decided to start it.
Atgenomix is a data and ML company that provides an open standard, unified platform for operationalizing biomedical data, analytics, and ML to deliver on the promise of precision medicine.
Atgenomix is a data and ML company that provides an open standard, unified platform for operationalizing biomedical data, analytics, and ML to deliver on the promise of precision medicine. Atgenomix envisions that cloud technology will transform healthcare in routine clinical genomics using artificial intelligence, and aims to provide the industry with modern computational and analytics technologies needed to scale data and ML in labs of the future.
"All happy families are alike; each unhappy family is unhappy in its own way."
Leo Tolstoy
Anna Karenina
My grandfather died of colon cancer when I was studying abroad and my wife was diagnosed with autoimmune diseases (antiphospholipid syndrome and lupus) after several miscarriages. A few years back, I learned about how next-generation genome sequencing can transform personalized medicine. I made a promise to myself that every family should be allowed to take steps to prevent disease. That is why I started Atgenomix.
Tell us about your SeqsLab Platform and how it helps customers.
Atgenomix SeqsLab cloud-native platform empowers clinical laboratories to build their own fully managed precision medicine production from sample to report with an integrated solution stack of cloud services, open-source software technologies, and industry standards, such as Global Alliance for Genomics and Health (GA4GH).
Building a production-grade workflow is like building a smart factory. In real-world systems, bioinformatics and ML algorithms are only a small fraction, while the required surrounding infrastructure is enormous and complex. One of the key considerations is workflow reproducibility which involves system setup, deployment, data collection, data processing, data analysis, and all the activities to automate and optimize the entire workflow as much as possible. Nowadays, developing workflow systems is relatively fast and cheap because of available open-source tools, but deploying and operating them over time is difficult and expensive.
We developed Atgenomix SeqsLab as a fully managed application that configures all the required technologies and resources to support and run your reproducible workflows.
Therefore, we developed Atgenomix SeqsLab as a fully managed application that configures all the required technologies and resources to support and run your reproducible workflows.
At the same time, you have full control and maintain ownership of your data and infrastructure. Our key strategy is to tailor and unify the required infrastructure with workflow design, not the other way around. We want to eliminate researchers spending 90% of their time and effort debugging and modifying workflows iteratively which is a compromise to fit into the execution environment.
Outcomes include generating insights faster and cheaper pre- and post-approval of drugs, supporting companion diagnostics programs, and providing real-world insights into genomic testing to support the biopharmaceutical market and commercial strategies, etc.
Why is ML so critical to data analytics and accelerating research in the future?
In the commentary article, “Precision Medicine in 2030 – seven ways to transform healthcare,” Joshua C. Denny and Francis S. Collins share, “Looking forward, biomedical datasets will become increasingly ready for analyses. [...] Applications of machine learning approaches could result in new taxonomies of disease through genomic, phenomic, and environmental predictors.”
One of my favorite tweets is from Francois Chollet: “Winners are those who went through *more iterations* of the ‘loop of progress’ – going from an idea, to its implementation, to actionable results. So the winning teams are simply those able to run through this loop *faster*.”
ML is a critical software tool to help us understand and build more complete models of diseases and go through the analysis iterations much faster to actionable results.
Therefore, this is the key reason we developed Atgenomix SeqsLab, which empowers researchers to achieve more, faster, by focusing on data analysis and machine learning, not the underlying hefty and complex infrastructure.
Atgenomix’s solution also works on analytical data from the lab in addition to clinical data. Atgenomix is focused on workflow automation, acceleration, and management of data processing and computing. Customers benefit from a unified, one-stop solution to manage multi-omics, multimodal data, and processing workflows.
What can we expect to see from Atgenomix in the coming months?
There are many exciting new platform capabilities coming in the following months. Users will be able to seamlessly mix SQL query commands with their automated and reproducible analysis workflows. The capability implements modern data lakehouse architecture and lets users distribute structured data processing using SQL queries and run 10x to 100x faster on existing workflows and data. We will also continue to deliver built-in ML capabilities that automatically run in distributed CPU or GPU computing clusters. These will further empower researchers to design and build scalable data analysis workflows with no limit and thus easily advance their biomedical research on diverse data at scale and speed.