« Pelosi Trip May Violate Separation of Church and State | Main | The Eugenics Trap »
May 30, 2007
NMPDR - Fighting Back at Drug-Resistant Diseases
Bacterial drug resistance is in the news today, thanks to a man infected with a drug-resistant form of tuberculosis who chose to expose hundreds of people to his condition.
The problem of drug resistance is due to a characteristic of bacteria known as stress response. One stress response is to suppress some of the chemicals that prevent errors in the DNA replication process, causing the organism to evolve more rapidly. In an environment full of antibiotics, the goal is to literally pump the medications out of the cell before they can do any damage. A tuberculosis bacterium that has built pumps for all the major anti-tuberculin drugs is called XDR-TB (eXtremely Drug Resistant). The unidentified man at the center of the current air travel contagion scare has XDR-TB.
The next-generation tools for fighting disease are drugs that turn off individual proteins in the bacterial cell. For example, one of the most promising anti-AIDS drugs disables the molecule HIV uses to protect itself from the immune system. The NMPDR web site (which is what Bruce works on when he's not working for me) provides tools for finding key proteins that can be targeted by new drugs. For example, you can get a list of genes in tuberculosis related to stress response. Once you've found a gene you want to target, the next step is to look for a matching protein molecule at the protein data bank web site. This gives you the shape of the protein you're trying to destroy and the electrical charge on its various surfaces. Computer models of promising drugs are then matched against the protein data to find one that will disable the protein and render the bacteria that rely on it harmless.
The matching is is a slow, difficult process that requires a lot of computing power. On the NMPDR project, they use a neural network to select drugs that are likely to work. This reduces the processing time to about one week per protein, but the drugs chosen are not necessarily the best ones, just likely good ones. Of course, identifying a possible drug is only the start. You need to run laboratory tests to verify that the computer predictions are correct, and then you must navigate the drug through the long process of FDA certification.
Still, it's much better than trial and error, and stories like those of the TB honeymooner bring into sharp relief the need for this kind of research.
Respectfully submitted,
Ferdinand T. Cat
# At Wed 11:57 PM | Permalink | Trackback URI | Comments (0) | More NMPDR | Tags: antibiotics bacteria bioinformatics biology drug resistance drugs in silico screening medicine stress response tuberculosis
Trackback Pings
Comments
| HTML is not allowed in comments; however, if you put in a raw URL (http://www.somewhere.com/page.html) it will automatically be converted to a link.. Also, it is likely your comment will not appear unless you refresh the page manually after posting it. |
