Overcoming Camera Blur

Model of Protein Folding and motion blur

Illustration of DNA Folding 

Image Credit
Steven Burrows

The basic question of how strands of nucleic acids (DNA and RNA) fold and hybridize has been studied thoroughly by biophysicists around the globe. In particular, there can be unexpected challenges in obtaining accurate kinetic data when studying the physics of how DNA and RNA fold and unfold at the single molecule level. One problem comes from temporal camera blur, as the cameras used to capture single photons emitted by these molecules do so in a finite time window that can blur the image and thereby skew the kinetics. In a paper published in the Journal of Physical Chemistry B, JILA Fellow David Nesbitt, and first author David Nicholson, propose an extremely simple yet broadly effective way to overcome this camera blur. According to Nicholson: "we wanted to measure the speed of nucleic acid folding, but to our dismay, we encountered a systematic bias that comes up when you do these kinetic measurements, if you're not careful. Specifically, you get into trouble, if the DNA is folding as quickly as the camera is recording images. So, we started thinking, how can we come up with a way to fix this problem?" In looking at their data, Nicholson and Nesbitt realized that they could reduce this systematic error and extend the domain of kinetic study in a surprisingly simple way by shining a light on the problem, in particular, a strobe light.

Why Throwing Data Away is Beneficial... Sometimes

Nicholson and Nesbitt realized that they could reduce the camera blur by shining pulsed rather than continuous laser light onto the molecules, reducing the time fraction of observation (duty cycle), but making their total error due to camera blur significantly lower. Nesbitt explained: "This was entirely David Nicholson’s idea, for which I simply gave him a little encouragement to pursue. The key idea is really like a stroboscope in a crowded night club or disco. Basically, we're flashing light onto the molecular 'dance' in a short time window faster than the camera frame rate. This of course requires us to throw away information between pulses, but at the same time, provides much better kinetic information from each pulse." The thrown away data can be accounted for by a mathematical correction, resulting in kinetics that are accurate even up to the camera frame acquisition rate, an order of magnitude improvement. Nesbitt clarified the fix in more technical terms: “The method reduces the error because normal analysis of blurred objects has a built-in mathematical bias that tends to make kinetic analysis of these actions appear systematically slower.” Nicholson and Nesbitt had just found a simple solution for their problem in a stroboscope.

A stroboscope is a fancy word for a strobe light. Stroboscopic imaging, a process where an objects movement is represented by short light samples, has been used for many decades, though not in single-molecule kinetic measurements. "Actually, we thought for sure someone would have already published this concept long ago," Nicholson commented, "but after digging through the literature, it turns out there was still a lot of confusion about how to deal with single-molecule camera blur. So, we said, 'OK, this seems like something researchers could find useful.’" The fact that this simple method had not already been published made Nicholson and Nesbitt more interested in making their method public for everyone to use. They realized that this method would help save researchers valuable time when it came to data analysis and to extend their kinetic measurements up to the camera frame acquisition rate limit. Nesbitt said: "I think comparing our stroboscopic method to a flash on a camera is helpful. We're taking images with this method to get rid of motion blur. It's allowing us to see our DNA molecules more crisply."

With regard to the simplicity of such a method, both Nicholson and Nesbitt initially wondered if this approach was even worth publishing at all. "Once you appreciate the basic idea behind the method, it seems so completely simple, and frankly, a bit obvious! We wondered,' is this really new and worth publishing?' But in fact, it's exactly those sorts of discoveries that belong in the literature because everyone can so easily implement it." As Nicholson added, this method would be inexpensive for researchers to implement, and they would see an immediate improvement in their range of measurement by as much as an order of magnitude. The team hoped that their method could save researchers time and effort when it came to fixing kinetic measurement bandwidth problems associated with camera blur.

Spreading the Word

Nesbitt and Nicholson look forward to seeing their work implemented by other researchers, "The wonderful corollary is that, as camera technologies get better and faster, David Nicholson’s method should improve right along with them," Nesbitt explained. The benefits of this new method are not only cost-effective and easy to use, but clearly can adapt as the technology itself improves in the coming years.

Written by Kenna Castleberry, JILA Science Communicator

Synopsis

The basic question of how strands of nucleic acids (DNA and RNA) fold and hybridize has been studied thoroughly by biophysicists around the globe. In particular, there can be unexpected challenges in obtaining accurate kinetic data when studying the physics of how DNA and RNA fold and unfold at the single molecule level. One problem comes from temporal camera blur, as the cameras used to capture single photons emitted by these molecules do so in a finite time window that can blur the image and thereby skew the kinetics. In a paper published in the Journal of Physical Chemistry B, JILA Fellow David Nesbitt, and first author David Nicholson, propose an extremely simple yet broadly effective way to overcome this camera blur. 

Principal Investigators