June 29, 2016

Electrons and X-rays working together, spurring new ideas?

Hans Elmlund and Andrew Martin are Imaging CoE AIs. One works on developing new image processing theories for single-particle cryo-EM, the other works in the realm of theoretical physics. One works at Monash, the other at Melbourne. One has a thing for electrons, the other for X-rays.

What happened when they both attended the Physics Symposium in December 2015?

Hans heard a talk by Andrew, then Andrew heard a talk by Hans, and now they are working together to enhance the quality and innovation of their respective research programmes. Hans and Andrew told us a little about their story, it’s a love affair involving big data, algorithms and software.

“I am slightly closer to the biological side of things than Andrew,” Hans begins. “But I was deeply involved in the first X-ray free electron laser experiments on biological particles at the Stanford Linear Accelerator Centre and when I arrived in Australia to work in the Imaging CoE I was excited to find out that our physics node had similar interests.” Andrew says that his work aims to realise the potential for X-rays to study the structure and dynamics of biological molecules.

“To do this requires new theory and data analysis methods. Our group studies the unique imaging situations that arise with very intense and very short duration pulses,” he says.

Both agree that single particle imaging with X-ray lasers and cryo-EM have a common goal: determining the structures of biological molecules.

The biggest difference, cryo-EM takes direct measurements of the phase while with X-rays all the phase information is lost and must be computational recovered.

“Andrew has been working a lot with understanding radiation damage and how it influences the metrics used to generate 3D images of biomolecules,” Hans explains. “In cryo-EM this is a largely unexplored area. My hope is that together our groups will be able to contribute to a better understanding of the implications of radiation damage associated with cryo-EM imaging,” he says.

“While our group can help with radiation issues, there is a great opportunity to use the algorithms Hans developed for cryo-EM to analyse our X-ray data,” Andrew counters.

“X-ray experiments measure diffracted intensity while cryo-EM experiments measure images or ‘take photos’. Our collaboration is based on identifying the implications of these different approaches and then figuring out how to adapt the data analysis methods. In the future, we hope the inverse will work, applying analysis ideas developed in the X-ray community on cryo-EM data.”

High-impact fundamental scientific discoveries take time, devotion, collaborative spirit and resources. The Imaging CoE is uniquely placed to facilitate interdisciplinary collaborations between Australian experts – this particular alliance is one of many we hope will develop over the lifetime of the Centre.

“By integrating information, data, techniques, tools, perspectives, concepts, and/or theories from two or more disciplines, we will be able to solve problems whose solutions are beyond the scope of a single discipline or area of research practice,” Andrew says.

While they don’t think this will make discoveries faster, they are confident that their cross-disciplinary interactions will open up routes to new discoveries and ideas that would otherwise never have come to exist.

“We are still in the early stages of extending the capability of Hans’ software for X-ray data,” Andrew says. “Once this is done we can test and compare our algorithms in parallel on X-ray and electron data.”

“I agree with Andrew,” Hans follows up, “the first goal is to provide a computational framework that can be applied to analyse any kind of projection image. To do this we will adopt techniques that have been developed for analysis of continuous diffraction patterns to the analysis of cryo-EM images and vice versa. Both our teams will continue to develop powerful new statistical algorithms for data analysis of large populations of noisy images of particles in different unknown orientations and with different structural composition/conformation.”

“We both hope in the future we will see synergetic approaches that combine information from both techniques to gain maximum understanding of the systems that we study,” they conclude in unison.

For further information about Hans’ open access software packages visit the SIMPLE website.

And, if you’re interested in being connected with other Imaging CoE researchers, not at your specific location, please get in touch with Stephanie Pradier or Manoj Sridhar. If you know who you would like to work with they can put you directly in touch. Or, if you’re not sure of what you need they can help advise of other researchers for you to work with.