Integrating 3D biological data on scales from molecules to cells

At the highest level of detail biology becomes a study of molecular machines. This is a truly complex and fascinating world; for instance, molecular machines generate forces, assemble or disassemble other molecules, function like channels, and store or transport energy.

Well established techniques such as X-ray crystallography, nuclear magnetic resonance spectroscopy (NMR) and electron microscopy (EM) are capable of visualizing the three-dimensional (3D) structures of these machines at the molecular level, but usually do so at the cost of studying them in isolation and in a state frozen in time. As with everyday machines, molecular machines may need to change state, change shape or move parts in order to perform their function. In analogy with an assembly line in a factory, many biological processes involve a multitude of molecular machines and a frozen glimpse of isolated machines is therefore insufficient for understanding the "big picture".

The emergence of new microscopy techniques such electron tomography (ET), soft X-ray tomography (SXT) and 3D scanning electron microscopy (3DSEM) now makes it possible to visualise molecular machinery in the context of cells, and even larger structures in 3D, albeit not yet at the level of detail afforded by the established techniques mentioned above. The challenge now is to integrate imagery at different microscopic scales to gain a broader perspective on the 3D micro-cosmos of biology.

In this project, we will take crucial steps in tackling this challenge by developing added-value services for the EMDB archive - the definitive archive for the 3DEM field managed by the Protein Data Bank in Europe (PDBe, located in Cambridge) and its international partners. We will achieve this by developing:

a) A web-based visualisation tool (volume browser) that provides a user with an easily accessible integrated view of 3D biological data at the cellular and molecular scales derived from the EMDB and PDB (the definitive archive for atomic coordinate models of biological macromolecules) archives, the means to navigate seamlessly between them and to link to other relevant bioinformatics resources;

b) A web-based tool for adding biological annotations to the archived data in EMDB, which is vital to achieve the integration described in (a);

c) A pilot image archive for "raw" experimental data that could be used for developing new EM methodology as well as for checking the validity of 3D structures archived in EMDB.

The web-based volume browser and the new archive will make a variety of 3D structural information vastly more accessible to the wider biomedical science community. The context and perspective gained by linking cellular and molecular structure data will provide unique insights that cannot be gained from the individual structures alone. Moreover, often these data sets contain much more information than what was exploited in the original publication. Therefore, the volume browser will become a vital instrument for knowledge discovery.

In this project we will develop two tools and a new archive that will add considerable value to EMDB data and enable its integration with other structural and bioinformatics resources:

a) Volume browser: We will develop a web-accessible tool to provide an integrated view of 3D biological data at the cellular and molecular scales derived from the EMDB and PDB archives, the means to navigate seamlessly between them and to navigate to related bioinformatics resources.

b) Segmentation-annotation tool: PDBe has developed a draft EMDB segmentation-file format (EMDB-SFF) that accommodates a range of segmentation representations used in the field as well as the requisite annotation. We will develop a web-based tool that will allow users to fully exploit the capabilities of this format by uploading segmentations in any common format, overlaying segmentations on the 3D volume, adding biological annotation, and then writing out the segmentations in the EMDB-SFF so that it can be included in an EMDB deposition.

c) Pilot image archive: Raw image data cannot be deposited in EMDB for a number of practical reasons - the size of single-particle datasets can exceed that of the deposited reconstructions by orders of magnitude, and there is no consensus as to what constitutes "raw" data. We will develop a pilot image archive for 2D image data related to EMDB entries, which will allow us to comprehensively address these issues in consultation with the EM community, provide public access to this data, and consider options for the future integration of the archive with other resources.

The pilot image archive will mainly impact the EM community. It will a) lessen the burden of image archiving for individual labs, b) be a valuable source of well-annotated data for methods development and testing, c) be a valuable resource for data that can be used for the validation of 3D reconstructions archived in EMDB, and d) will provide a rich source of well-documented test cases that can be used in teaching and training of researchers entering the EM and ET fields. Resulting improvements in the quality of EMDB data will promote its wider use in biomedical research.

The Segmentation-annotation tool will enable the deposition of biologically annotated segmentations to EMDB, thereby substantially enhancing the potential for integration of EMDB data with data from the PDB and other bioinformatics resources. In addition to benefitting the EM community, this web service may also be used by related imaging fields that use segmentations to interpret 3D volumes, such as the biological soft X-ray tomography and 3D scanning EM communities. Improved integration between EMDB and other bioinformatics resources will also stimulate use of or reference to EMDB data by these and other resources. As many biomedical researchers will be more familiar using these resources, this will substantially increase the exposure of EMDB data to a wider biomedical audience.

The Volume browser tool will be a unique publicly accessible resource providing an integrated view of 3D data from EMDB and PDB -as well as data from relevant bioinformatics resources- spanning scales from cells to atoms. For example, tomographic reconstructions with overlaid segmentations will provide the 3D cellular context for biomacromolecules to molecularly oriented biologists, whereas scientists approaching their problems at the cellular level will be able to find high-resolution structures of molecules associated with segmented regions of tomograms. The presentation of data in an easily accessible manner and the enhanced integration of data will create opportunities for the discovery of new knowledge by the wider biomedical community. For teachers and students of biology and biochemistry, the resource will provide an evidence-based visual exposition of many key biological systems and processes. For the pharmaceutical industry the resource will open new avenues to a visual understanding of how potential drug targets behave in the context of their cellular environment. Reciprocal links from bioinformatics resources to the volume browser will improve its ranking in common search engines such as Google when searching for general biological and medical terms used by the public. It will thus provide a valuable entry point for general users to develop a better visual appreciation for the basis of biology and medicine on the microscale.