Already in 1968 one recognized that the transmission electron microscope could be used in a tomographic setting as a tool for structure determination of macromolecules in a biological sample.
This technique, commonly refereed to as Electron Tomography (ET), has a unique advantage: It is currently the only technique that allows one to study the structure of individual molecules in an amorphous biological specimen, something which is important in order to address several key biological issues in life sciences research. However, in order to truly capitalize on this advantage, one must be able to reliably recover the structure with a resolution that is high enough.
Unfortunately, due to a number of mathematical and experimental problems, it has been difficult to obtain reliable reconstructions unless sophisticated reconstruction methods are employed, and such methods are not routinely used within the three-dimensional electron microscopy community. This in turn has lead to a situation where most applications of ET in structural biology are confined to the study of cellular substructures rather than studying individual small molecules (''small'' in this context refers to protein molecules of size smaller than 200kDa). Moreover, ET is still considered as an emerging technology within structural biology.
The above mentioned experimental and mathematical problems have been (and still are) responsible for the slow dissemination of ET as a reliable structure determination technique in life sciences. The former issue is partly addressed by the rapid technological development in sample preparation and instrumentation, and currently state-of-the-art instrumentation allows one to routinely collect high quality tilt series that can be used in ET. Unfortunately, the same can not be said for the development of mathematical methods and algorithms that addresses those issues that are related to the high noise level in the data and the severe instability (ill-posendess) of the reconstruction problem. Hence, to utilise ET to its full potential, one needs not only to have access to good instrumentation, but also make full use of the physics models and mathematics used in the structure recovery step.
To make full use of the physics models and mathematics requires an highly interdisciplinary scientific research effort. It turns out that there are a number of open mathematical research problems that needs to be addressed. These are of both numerical as well as theoretical (conceptual) nature. This involves expertise from the mathematics of tomography (integral geometry), the mathematics of stable and efficient computations (numerical analysis, regularisation theory, and harmonic analysis). Moreover, a careful modelling of the physics of electron-specimen interaction and the optics in the microscope is required to be able to recover structures of molecules at high resolution. This in turn involves expertise from quantum scattering theory and quantum wave optics. Furthermore one needs expertise in software development for large scale scientific computations and visualisation as well as image processing. Finally, if one seeks to obtain the highest possible resolution while maintaining the reliability, the methodological developments mentioned above all need to account for part of the prior knowledge about the sample from structural biology (e.g. antibody structure and dynamics), cell biology, and bio-informatics.
It is rather clear from the previous paragraph that it is highly undesirable to attempt at pursuing such an interdisciplinary research effort in-house. Therefore, Sidec believes that methodological development in ET is s best pursued with the academic community in an open fashion. Therefore, these academic collaborations do not focus on the commercial side of Sidec's activities, but rather the focus is on developing the necessary mathematics, algorithms, models from physics and software components in order to make optimal usage of ET as a reliable structure det method within structural biology.