IOMICS Corporation is an award winning analytics company based in Worcester and Cambridge Massachusetts. In April 2016, IOMICS announced the release of its FUSION Analytics Platform™, a cloud-based software system for prescriptive analytics and rapid prototyping of advanced decision models for use in chemical engineering, medical research, and clinical care. FUSION is hosted at the Massachusetts Green High Performance Computing Center (MGHPCC) under aegis of the Commonwealth Computational Cloud for Data Driven Biology (C3DDB).
C3DDB is an MGHPCC-based initiative connecting life sciences research with emerging, innovative analytics companies like IOMICS. C3DDB is open to groups conducting life sciences research or advanced product development at academic institutions, research institutes, and companies throughout Massachusetts.
The IOMICS leadership team includes CTO and Senior Software Architect, Joseph Gormley; CSO and Computer Scientist, Daniel Corkill, Ph.D.; and Software Engineer and Technical Lead, Thomas Zisk.
“The pace and complexity of life sciences research is increasing rapidly. The field has become data rich, yet cost-effectively obtaining actionable knowledge remains a challenge,” Gormley says. “The problem goes far beyond just building a data warehouse or using the latest visualization tools. To maintain progress, academic and industrial researchers must be willing to embrace entirely new computational models capable of extracting meaning from a broader range of multiscale clinical and molecular data types.”
“The FUSION Analytics Platform utilizes the most effective practices in data science while automating many of the manual, tedious and error-prone steps in data staging, prescriptive model building, model validation and deployment,” Corkill says. “One of the most innovative features in the platform is its ability to continually assess and improve model performance under changing conditions and as new data and supporting information become available. We want FUSION to guide users in following best practices and to alert them when there are reasons for concern. We also strive to allow FUSION to make the best use of whatever usable data are available—especially in cases where scarce data has to be utilized carefully and fully in developing precise, targeted models.”
“FUSION is not a traditional desktop tool or pipeline product,” Zisk adds. “We worked hard to encapsulate many of the most important aspects of predictive and prescriptive model building into a streamlined cloud-deployed service. Our user interface permits dataset development, model build, model validation, and model deployment without ever having to leave the platform.”
The FUSION Analytics Platform provides the latest machine learning techniques, including multiple deep learning strategies, with over eighty open access biochemical and omic data sources. The FUSION Analytics Platform excels at automating data-driven model development utilizing a diverse range of semantically integrated biochemical and molecular data sources including some of the most used but difficult to manage datasets from TCGA and Pubchem. FUSION is available as a general development platform and currently used within three application verticals: FUSION STS™ for chemical screening, FUSION BD™ for biomarker research, and FUSION CI™ for population health. All FUSION models can be used standalone or as microservices embedded within other products and services including next generation discovery workflows and clinical decision support systems.
A Fortune 500 Corporation’s Advanced Materials Division utilizes FUSION STS for high throughput screening of chemical compounds. Their goal is to reduce or eliminate the hazardous properties within consumer food packaging. “If we want a healthier planet we need to shift our thinking away from only downstream regulatory testing to much earlier and continuous toxicity screening for all product development initiatives,” Gormley says. “Our expectations for safer products are much higher today than they were just a decade ago. Part of the continuing problem, however, is the product development paradigm itself. Chemists and chemical engineers have unintentionally designed products with inherent safety problems due to the lack of reference data and analytical tools to model and characterize a broad range of chemical hazards within food, packaging materials, cosmetics, pesticides, and more recently, nanomaterials. Consumers have expected downstream regulatory testing, with all of its inadequacies, to keep us safe for a given level of exposure. That approach to chemical testing will continue to allow potentially dangerous products into the marketplace. We believe the solution is to apply safe-design practices continuously and much earlier in the product development cycle. Our FUSION STS application promotes safer consumer product design through early and continuous in silico toxicity screening.”
FUSION BD, a second IOMICS vertical application, has been equally well received. Researchers at WPI and Harvard are planning to utilize FUSION BD for multiple applications from automated mining of cancer data sets from TCGA and internally generated experiments, to automated monitoring and control of lab-on-a-chip processes. Researchers worldwide have long been seeking powerful yet simple-to-use tools for multiscale data staging and data-driven model development in support of biomarker discovery. FUSION BD allows researchers to identify clinical and molecular signatures of pathogenic processes and disease progression for new drug development and companion diagnostics, to support clinical trial monitoring, and for improved post-marketing safety assessment.
FUSION CI is the the third and newest application to be produced from the platform. Regional healthcare providers are evaluating its use in identifying rising-risk patients within clinical populations. FUSION CI is unique in its ability to use multiscale chemical and clinical data types ranging from clinical observations, to established PRO instruments, to high-complexity assays. The goal is to build precision models that identify and rank interventions based on current patient state and historical population events leading to actionable care improvement strategies. IOMICS will be presenting this work at the 2017 Bridging Practice and Theory Summit (BPTS) sponsored by Drexel University.
“FUSION is about automating complex knowledge discovery. Each of these vertical applications are examples of how precision models, when developed on the right analytics platform, can be rapidly built, evaluated, and placed into service, as cloud-based applications accessed from your desktop or as real-time microservices, to increase research quality and accelerate the development of critically needed new products,” stated Gormley. “There is enormous potential for data science and prescriptive analytics to drive improvements in our world – we are excited to be contributing to this rapidly developing field.”
IOMICS Corporation is an analytics company that believes the freedom to explore new ideas and rapidly deploy advanced analytics is essential to accelerating science, designing safer consumer products, and improving healthcare. Our mission is to put advanced decision models to work everywhere—in the lab, in the clinic, and in the environment. IOMICS encourages collaborative model development and co-branding of turnkey applications for chemical engineering, medical research, and clinical care. Additional company and product detail can be found at www.iomics.us.
Story image: IOMICS was recently named to Drexel and CIO.com Analytics 50
