Racing to decode the genome for malaria, which kills more people each year than cancer, Assistant Professor Joseph DeRisi, is constantly searching for ways to streamline laboratory processes. Depositing DNA material on slides by hand, for example, is an excruciatingly slow procedure. "Working manually, it can take a day to study a single gene. But by automating the process, we can study anywhere from 6,000 to 40,000 genes in the same time frame," says DeRisi, head of the University of California at San Francisco's (UCSF) DeRisi Laboratory.
So the mechanically-oriented assistant professor of biochemistry and biophysics opted to build his own robotic device. Called a DNA microarrayer, the 3-axis robot precisely applies tiny droplets containing DNA to 250 glass slides simultaneously. Linear motors from the Compumotor Division of Parker Hannifin (Rohnert, Park, CA) provide the high accelerations (up to 5 g's) and speeds required. Motor controllers from Galil Motion Control (Rocklin, CA) deliver the pinpoint accuracy and repeatability.
An earlier version of the microarrayer used ball screw technology. "Although it served its purpose in research, it was slow, and backlash was a problem," DeRisi says. "The new microarrayer has no backlash, and the linear encoder on the motor gives excellent repeatability to within one micron."
In action, the robot uses 48 specialized printing tips to pick up a DNA-containing compound, hold it tightly, and print many slides in precisely the same place. It then travels to a station where a sonicator cleans the tips and a vacuum pump dries them before moving back to the source tray to pick up the next sample.
The positional accuracy is plus or minus 2.0 microns, achieved by the controller, which handles a very fast servo loop update rate. It provides information about tip position every 250 microseconds, or the equivalent of four update loops per millisecond. Each pick and place operation of the robot takes about 200 milliseconds.
Cost played a key role in the selection of motion components. Academic institutions that perform genetic research don't have the large amounts of money that commercial organizations funded by investors do. Backed by the National Institutes of Health (NIH), DeRisi not only developed the DNA micro-arrayer now in use at UCSF, he made it available to colleagues at other univer-sities by posting the building instructions and programming software on his website (http://derisilab.ucsf.edu/).
The x and z-axis stages of the DeRisi DNA microarrayer move in a highly coreographed fashion, as shown in this motion profile. The x-axis stage travels one inch in about 2.10 milliseconds and then stops. The z-axis print head travels 300 microns to tap the printing tips onto the glass slides. It immediately retracts to its reaised position and the whole process repeats.
Galil supplies a software development kit that DeRisi used to write the software that instructs the machine using C++, rather than Visual Basic.
DeRisi's research focuses on malaria. "We're working on decoding the malaria genome," he says. "About 2.1 million children die of it per year, and 300 million become infected. The only other organization that researches malaria using DNA chips is the U.S. military. We now have a malaria chip similar to the first organism for which the genome was decod-ed-brewer's yeast. We can use the malaria chip to identify drug responsive and drug resistant genes. For me, this is all about how to use robots and motion control to fight a very old disease."
|Performance specifications of DeRisi DNA microarrayer|
|Number of stages||3|
|Maximum velocity||3 meters/second|
|Maximum acceleration||5 g's|
|Peak force||330 Newtons|
|Repeatability||plus or minus 2.0 microns|
|Maximum throughput||14,400 spots/minute|