Here's an update on the Human Proteome Folding project from the Institute for Systems Biology (ISB), who owns the project.

Overview:

World Community Grid has sent large batches of results back to ISB and ISB is running them through the post-processing stage of the process, and the results look good. This means that World Community Grid is working just as it should. Our best estimates at this time are that Human Proteome Folding is from 15 to 20 percent complete. ISB will be delivering real information out to biologists within 3 to 6 months (the delay being the post-processing and human interpretation of a few of the results).

Validation:

All internal tests indicate that Rosetta is running correctly, that we're making good structures, and that we're not corrupting the input or output files anywhere. Several proteins have been run through the model-selection phase and show that the grid is operating correctly.

General Progress / General Results:

As of January 21, 2005, World Community Grid has processed approximately 9% of the protein sequences in the entire project. Grid.org is probably close to processing a comparable number. For each batch of 1,000 sequences, the work is subdivided into about 50,000 (+-20,000) individual work units. A heuristic, based on the length of a gene sequence, is used to decide how much to subdivide the work to make the work units average about 10 hours of run time. Even with this, the actual run-times vary over about a 10 to 1 range or more. Each work unit is sent to more than one device so that results can be compared to eliminate those that may have encountered some sort of problem. Because of the wide range of device characteristics and run-time patterns, the results arrive in a bell curve-like manner over time. Furthermore, because of the unpredictably long running work units, it takes extra time to receive the longer running work unit results for a batch of work. This leads to some imprecision about how far along we are.

ISB is starting the post-processing phase and will soon be populating a data-base of structures that biologists will actually use. So another way to state our progress is that we've accumulated enough results to start the data-analysis/post-processing phase.

Biological Results:

After the structures are preprocessed, scientists will compare these with proteins of known function to narrow their studies to the most likely candidate functions. The shape information will be used to look for interactions that may occur with other proteins, disease pathogens, etc. Currently our post-processing is revealing the shapes of these proteins. Images of interesting examples will be posted at the ISB site after some time. It is still too early in the process to tell whether we have enough information for curing specific diseases. However, the scientists at ISB are very positive and upbeat about the potential of this project.

Going forward:

In late February we plan to create a Forum specifically dedicated to questions and discussions about the Human Proteome Folding project. This forum will be monitored by scientists from ISB.

For more details about Human Proteome Folding, please refer to the following:

https://secure.worldcommunitygrid.org/projects_showcase/human_proteome.html
http://www.systemsbiology.org/Default.aspx?pagename=humanproteome