Scientific breakthroughs are built on insights from good data, but sometimes those breakthroughs are hidden under an ocean of too much of it.
Integrating data
Recently the California Institute for Regenerative Medicine (CIRM) approved a new kind of investment. A new infrastructure award aims to create tools to make sense of the massive, complex biological datasets produced through CIRM-funded research. The program would fund the creation of new software to integrate regenerative medicine data and make generate new insights.
Scientific breakthroughs are built on insights from good data. However, sometimes those breakthroughs are hidden under an ocean of too much of it.
Recently the California Institute for Regenerative Medicine (CIRM) approved a new kind of investment that would create tools to make sense of the massive, complex biological datasets produced through CIRM-funded programs and in the broader research community. The program would fund the creation of new software to integrate regenerative medicine data and make generate new insights.
“Our mission drives us,” Janie Byrum, CIRM Sr. Science Officer said during a board meeting this week. “We are designing tools and strategies to maximize the value of CIRM’s investments by advancing science and creating new ways to use data.”
Accelerate discovery
The new program would allocate $10 million to support 15 to 20 awards of up to $500,000 each to develop software solutions for integrating these massive datasets. Each program would focus on solving a specific data integration problem. Some will create new tools to connect different types of data. Others will improve or maintain widely used open-sourced software that works better for regenerative medicine. Some projects will update existing tools to make them compatible with modern computing and AI.
Called Data Science and Software Engineering Awards the goal is to create open-source software tools that would integrate different types of biomedical data. Connecting this data in a usable way could accelerate the identification and validation of new drug targets and biomarkers to treat disease.
In other words, the goal is to build the infrastructure researchers need to move faster and work more effectively across datasets. Focusing on software to translate the variety of types of data generated in regenerative medicine research and clinical trials will help accelerate discovery and make the data more easily shared.
A few years ago, the NIH launched a similar effort, called the Biomedical Data Translator program to integrate the many petabytes of biomedical data generated from federally funded studies.
For CIRM, this is also a way to build on its earlier investments in data sharing Data Explorer tool. The tool now includes more than 800 CIRM datasets, making easier to find and accessible, and easier for researchers to use.
See behind the mountains of data
Making sense of complex data has long challenged government‑funded research.
In 2014, former NIH Director Francis Collins warned Congress that the growing wave of data could overwhelm researchers.
“The challenge,” Collins said, “is how to store, retrieve, integrate, and analyze this mountain of complex data.”
Since then, the scale and complexity of data have exploded. Federal policy now treats data management as a core research responsibility. CIRM’s funded projects and clinical trials alone have produced more than 1,200 datasets. That increase strengthened the case for software tools that connect shared datasets. Advances in machine learning and AI now offer even more ways to tackle the problem.
Some tools already combine different kinds of biomedical data. However, many open‑source systems still need updates to keep pace with modern computing and AI. New or improved software could better meet the needs of regenerative medicine researchers.
The goal of integrating different data types is to speed up regenerative medicine research and maximize the value of CIRM’s initial investment, Byrum said.
Open, transparent, reusable, and extensible software could provide the infrastructure needed to overcome these barriers.
What’s next
CIRM’s new data‑sharing awards put software development front and center. Applicants must think not only about building a tool, but also how it will be deployed, used, maintained, and improved over time.
CIRM is looking for projects with broad impact. The agency wants tools that work across systems, data types, and research settings. The goal, make it easier to turn complex, fragmented data into meaningful discoveries.
Data science evolves quickly, and even widely used tools need regular updates to keep pace with advances in computing and AI. This investment helps researchers keep up with those advances and the continually changing landscape of new research data.