The following letters, written by Dr. Jonathan Rothberg, were posted to the Team Discovery forum on 29 & 30 March 2002. This is a time when the team is trying to determine its immediate future.
They are a great introduction to what the TSC project has as its goals, and an explanation of their intentions:
Thanks for your help.
This project is the closest to my heart.
We have the resources and long term commitment to make this the ideal team computing project.
I hope you can move the entire user base (you need some competition) from the UD cancer project as they approach the end of the project.
We want to keep the 1) Computing, 2) Scientific, 3) Cancer, and 4) TSC oriented volunteers motivated by addressing each of their needs. This is truly a team effort and we need all of you.
1) For the computing folks we want to make sure we improve the quality of this effort, including the look, feel, excitement and reporting. And... are looking for anyone with computing skills that want to help pitch-in in improving the tools, site, reporting, or porting to other systems like Unix and MacOS X.
We also know you want real results, and this is the only project with exceptional targets, control experiments done to make sure it is not a video game, real testable candidate drug compounds and an institute and funding dedicated 100% to testing and developing drugs based on the results.
2) For those driven by the biological science we are putting a huge effort into the elucidation of the pathways underlying TSC and the identification of the best targets (see Rothberg Courage Awards, with labs now going at Harvard, Yale, Fox Chase).
For those driven by the computational chemistry, and physics we are going to sponsor a $1,000,000 prize. We also are unique in that we are sharing our results in real time to make sure everyone can use them, and updating our discoveries on the pathways as fast as we can post them. We are also exploiting the best in genomics, computational chemistry, and informatics, and have created the first non-profit "chemical genomics" institute.
In molecular biology, computer science, and chemistry we are looking for full time people to come join the institute, and part time help.
3) For those that care about cancer, and this is a big statement, we feel that the focus of this program on the BEST science will have potentially a huge direct impact on the understanding of cancer and cancer drugs. Both because TSC causes many pre-cancers in many different organs (and some cancers) and because we are following up and testing all the drugs we get both on TSC and on cancers.
So if you care about breast, brain, lung, pancreatic, kidney, blood, or other cancers you should know that all cancers, and there are more then 200, share a small set of underlying "mechanisms" that they use to kill you. Our work on TSC is designed to understand this fundamental basis, and cut through this noise that has prevented the war on cancer from ending successfully.
4) For those that care about TSC. I am privileged to have seen the light on the end of the tunnel that technologies coming out of the genomics revolution, the computing revolution, and the chemistry revolution can produce. With focus we will make a difference, as a team we will have the resources to solve this problem.
We want to make drugs to help children with TSC. We would like to cure cancer.
Distributed computing will be a key to our success as will improvements in computational chemistry. We will insure this is not a video game.
It is our intention to do the best research in the world to find key novel drug targets for TSC. They will come in many classes of druggable target. So we will need many types of docking software, compare results from different systems, and encourage improvements. And, again, of most importance we use real compounds that we order and test in real TSC and cancer cells to determine which should be refined into drugs. We also make all the information available in real-time to the world as well as to our collaborators at Yale, Harvard, and the Fox Chase Cancer center.
I am in a privileged position and have had the first glimpse of all the potential drug targets from the human genome. There are 8,000 of them. Only a handful will lead to drugs for major diseases (currently the drugs are made against targets encoded by 350 human genes). And even a smaller number will be targets for cancer drugs that do not kill cells in general (that is how almost all cancer drugs work).
1) I can promise you communityTSC will have the best drug targets for TSC and cancer. Period. (Our advisors are the best in world at finding drug targets, and I'm pretty good at it too )
2) We will use the best algorithms and test them. And pay to improve. We are committed to have the best docking based VLS systems running (e.g. while I can't pick the best wine I know people that can). We will keep people informed on progress and improve the visualization tools and statistics to make it interesting to participate.
3) We will share all results in real time for anyone to follow-up, as well as continue to sponsor labs at Yale, Harvard, Fox Chase and others to get new targets and study our results. (We have a good track record here, I was involved in the first whole genome proteomics project and we published on the cover of Nature, where Watson and Crick published, and shared ALL our results on a web site).
4) We will work with the National Cancer Institute and the TS Alliance to conduct clinical trials. (I work with the guy that made two of the top selling drugs in the world, including Lipitor, and the former head of the NCI who cured Hodgkin's lymphoma and set up the drug screens that found Taxol, so we get some good advice).
5) My brother drives a new Ferrari and if you visit Boca and work on this project he will show you the time of your life.
We plan on identifying the key targets that will be useful for TSC. We do not just go to the literature and look where everyone else is looking (known as the "public targets". We use the best genomics, proteomics and bioinformatics available. Each new target in turn will have its x-ray structure completed, and will be distributed by communityTSC for people to help make drugs against. Some will be targets that may be central to a range of cancers including brain, lung, kidney, and others.
No one docking software is universal, in fact if you benchmark (test with known binding candidates) the major packages you will find that each works well on some class or "shape" of targets, enriching 20 to 50 fold from a candidate library. So if you have a million compounds you can get it down to 20,000 or so that you can test. Since you still need to test so many to get the real ones that bind, and it could cost $10,000 to make one compound we use real existing compound libraries so we can order the set of hits for $20 to $60 dollars each (or free from the NCI, which is the library we are using first) and test them.
Since no one "docking" package is best, we will distribute not only the current docking system we are using but others like DOCK, and we hope to test THINK. In addition since our ultimate goal is an enrichment of 1000 fold, we will be offering $1,000,000 for improvements to docking based VLS that achieve this enrichment over a range of target classes. Please any members that are good programmers help us port our current docking software to Linux and OS X. If you want to help we will put you up in Florida to do the ports.
Large pharmaceutical companies will focus on the big diseases and big cancers, and the big targets. Public docking projects use the publicly available targets which have actually all been looked at already by big pharma in both computational and "wet" screens.
We are focusing on elucidating NEW targets for TSC, an orphan disease, and related novel cancer targets. Big companies for the most part do not work on orphan diseases for obvious reasons (I know, I have raised over $600 million to fund my company and have over $1 billion committed from larger companies to make "blockbuster" drugs against novel "non-public" targets. I understand how the industry works).
By focusing on a simple system, such as TSC, using real drug-like candidates, encouraging improvements in computational chemistry, and testing all our results we will have an impact on TSC, create a foundation to do research for other orphan diseases (that we will freely share since a computational approach is the only economic way to go for them), and be able to do good science that will lead to a more strategic "war on cancer". The history of medicine is on our side, polio and tuberculosis, were all cured by people who took a different and more focused approach.
I hope you help.
by KeysCapt edited by lilhurricane
last modified: 2007-11-11 20:39:58