September 2003
« August 2003 | Main | November 2003 »
September 24, 2003
September 2003 Newsletter
- Comments on technical progress
- Comments on Goals
We will attempt to put out short newsletters about every 2-3 weeks. Originally we had hoped to give comprehensive updates, but too much is happening. Hence, we will just cover a few topics each time.
Comments on technical progress
We are rapidly reaching the point where we will have a functional version of the FIG database and software. The initial version will be called "The SEED", and it now contains 237 complete or near complete genomes; of these 72 have restrictions on use. Our schedule for making the system available is roughly as follows:
- As soon as we get one of the new dual-processor G5 Macintoshes from Apple (we would guess this will occur in 4-5 weeks), we will put up a version of the system on our server. It will contain just the 165 distributable genomes (this number may grow by a few before the actual release).
- The system is intended to act as a framework for community annotations. Hence, we will make it freely available to any sequencing project that wants to use it. We will provide detailed instructions on how to install local copies of the SEED, and groups will be able to add their own genomes. We plan on initiating this service in November.
- We will hold a class jointly with Argonne National Lab on how to use the SEED, as well as tools being developed at Argonne, to support community annotation of genomes. It will be held in late November. This will be a 2-day class. If you wish to participate, please let us know, since we may have to limit the size (it will be a "hands-on" class, so we want to hold the size down). There will be no charge and no support, and anyone that wants to annotate genomes will be welcome up to the size limit -- which has not yet been fixed.
- We will set up the capability of downloading copies of the system over the network, along with detailed instructions for installation. Those institutions with good bandwidth will wish to use this approach.
- For individuals or institutions that cannot download gigabytes over the network, we will provide "FIG disks". Initially these will be for Mac OS X. Veronika, Andrei Osterman, and I all run the system on our G4 laptops, and I anticipate that Gary Olsen and Rob Edwards will use this platform as well. Niels Larsen and Ralph Butler will go with Linux, I think; so that means that Mac and Linux disks will be available. This means basically that anyone can spend $250 for a disk drive and put the SEED up on either a Mac or Linux system.
Comments on Goals
The goals of FIG are not very precise; they differ between participants and we do not seek consistency (just good will and pleasant discussions). Last week Ross got into discussions which have stimulated some argument. Here is the position he took:
- There will be about 1000 complete (or very near complete) genomes within about 3 years.
- We should consider the problem of providing the best possible annotations for these 1000 genomes by the end of the 3-year period.
- It might seem that this is obviously what the communities are trying to provide, however this is simply not true. Research biologists are rewarded for making substantial advances on hard problems, while a great deal of the task of providing "best possible" annotations relates to relatively mundane issues.
- The act of focusing on this single, relatively technical issue is at the heart of an important paradigm shift -- the move from focused research to "high-throughput annotations" as a vehicle for driving an overall advance.
The claim is that, if we attempted to engineer a solution to the problem of producing the best possible annotations for all 1000 genomes within the time frame, the steps that would be taken will actually be substantially different than those that are now envisioned.
Ross is considering casting this as a concrete goal of FIG (to make the goal precise, to produce a detailed plan to achieve it, subject it to criticism, and then gather participants). If this strikes a cord, please feel free to communicate directly with him on the issue (Ross@TheFIG.info). If you are offended by such a broad focus, rather than attempting to isolate the significant issues and focusing on them, also feel free to comment.
We will send out short newsletters more frequently just to keep people informed. If you would prefer to be dropped from the mailing list, please let us know.


