Cell engineering can give IBM Watson “microscopic eyes”
What if your cells could actually see? Today, scientists at IBM Research’s Almaden lab in the Silicon Valley are pioneering a new field of discipline – cell engineering – that uses cells as sensors to transmit new information to researchers about disease and the environment.
In collaboration with University of California-San Francisco, our lab will play a major role in the new Center for Cellular Construction, an initiative that the National Science Foundation has funded to the tune of $24 million. The grant also includes researchers from San Francisco State University, Stanford University, UC Berkeley and the Exploratorium science museum.
“Our collaboration will produce an unprecedented amount of data,” said Simone Bianco, research staff member, IBM Research – Almaden. “Using cells as sensors, we can develop cognitive maps that help us understand the relationships between cell structures and functions that IBM Watson can further analyze. In a sense, we can use cells to give Watson ‘microscopic eyes’ so we can better understand cellular behavior in different conditions, from complex environments to human diseases.”
With “microscopic eyes,” IBM Watson can analyze hundreds of thousands of microscopic images to help accelerate the pace of research in biology. Cell engineering combines several existing disciplines — from physics and biochemistry, to cell biology, mathematics and computer science – and relies heavily on massive amounts of data. Watson can help uncover relevant, sometimes “invisible,” features of the data with unprecedented speed and accuracy.
“If we understand how cells structure themselves in normal conditions, we can use this information to infer any abnormal state,” said Bianco. “For example, the exposure of a bacterium to a toxin may cause it to shrink, while exposing it to a gas may make it bigger. Without even knowing what causes the changes, and with our machine learning algorithms, we now have a way to better study abnormal conditions. The information can help with a number of applications, from uncovering new advances in disease therapeutics to increases in efficiency of biochemical reactions.”
IBM Research will drive several aspects of the new center including:
Computer Aided Design (CAD): Researchers aim to create a computational platform — modeled on the CAD software used by mechanical engineers — to aid researchers around the world in precisely and predictably designing cells and multicellular structures with desired functions based on knowledge of their inner workings.
Cell State Inference Engine: The center will design image analysis software to enable researchers to use cells as living sensors of environmental conditions — with real-world applications such as monitoring air pollution or serving as industrial quality control sensors.
“It is another unusual feature of our center to have a major industrial partner as a core member of the center itself,” said Wallace Marshall, PhD, a professor of biochemistry and biophysics at UCSF who will serve as director of the new center. “We feel it is an important component for getting knowledge into the real world, while being able to take advantage of the major strengths of IBM in big data analysis.”