|5 Jun 2013|
Dr. Igor Jurisica explained the outcome of the Help Conquer Cancer project, and detailed the real-world implications that these results can have for cancer research.
Over the past five and a half years, World Community Grid volunteers on the Help Conquer Cancer project have analyzed almost 120 million images of protein crystallization trials. On May 29th, Dr. Igor Jurisica from the Ontario Cancer Institute presented an overview and summary of the recently-completed project, and explained the real-world implications that these results can have for cancer researchers everywhere.
Here is the replay of the webcast and question & answer session - featuring lead World Community Grid scientist Viktors Berstis and Dr. Igor Jurisica:
Despite remarkable progress in the past decades, cancer remains a significant health threat worldwide. Part of the global effort to diagnose, understand and treat it entails understanding the operation of certain key proteins that are involved in one or more cancers. However, it is quite difficult to perform large-scale analyses of these types of proteins for the following reasons:
- In order for the function of a protein to be understood, its physical form must be modeled very accurately.
- In order for a protein’s form to be modeled, it needs to be imaged and analyzed, most commonly by X-ray crystallography.
- In order for this imaging to be possible, the protein must be crystallized.
- Protein crystallization is fiendishly difficult, and depends on hundreds of factors, including the purity of proteins, super saturation, temperature, pH, time, ionic strength, purity of chemicals, and the volume and geometry of samples. Each particular protein the researchers need to find the specific conditions under which it will crystallize, and at present there is no way to predict what those conditions will be for each protein.
As part of an ongoing effort to understand the process of protein crystallization, researchers at the Hauptman-Woodward Medical Research Institute created a High Throughput protein crystallization screen, and researchers at labs around the world have been subjecting a wide range of proteins – over 13,000 so far – to a specific set of different crystallization precursors: 1,536 different combinations per protein. The results of each crystallization trial were photographed and stored, resulting in a huge image backlog: 120 million images that needed to be analyzed. Each image represents the result of a crystallization attempt for a certain protein under a particular condition and at a specific time.
This was the enormity of the task at hand for the Help Conquer Cancer project: to analyze each of these 120 million images. Only a small fraction of these were characterized by human experts, and many of the 13,000 proteins had no indication of successful crystallization.
Thanks to the generosity of World Community Grid members, the analyses of these images is now complete. The researchers will now compile the results, with the aim of advancing multiple related goals:
- Continue to increase the speed of the image-scanning algorithms and use the false-positive and false-negative results to further refine the algorithms to increase correct classifications of the images.
- Improve understanding of crystallization principles, to increase the number of successful protein crystallizations so that more protein structures can be determined and their function characterized.
The full project results of their analysis will be made available at a future date for use by other participating labs.
Other cancer researchers will use the valuable protein structure information to better understand the respective cancer processes, which should eventually lead to better treatments.
The researchers want to thank you, the many volunteers who have donated computing time to this project. The analyses would have been absolutely impossible without the amazing computing resources available for free through the World Community Grid.
Our work doesn’t end here! Stay tuned for new projects, including a new one from the Ontario Cancer Institute that will be announced later this summer.