It's a good question which I've answered earlier this year on the ARQL thread. The concept is difficult to explain because to a certain extent it's wrapped up in the world of biophysics and chemistry and their connection to human pathology.
Bottoms up refers to attempting to identify a chemical or cellular functional agent that on its own without hurting the host at the sub-cellular or cellular level to engage in disease destruction. Every individual has a unique genetic makeup so specificity in drug design commands a premium. Drug non-genetic specificity throughout the 20th century has had a questionable history and may only be nothing more than placebo. Some drugs especially those created to destroy disease mechanisms at the super-cellular, or organism, level are remarkably effective. It is the viral machines against which we have made little progress. They have the long term adaptability built in, so even when we can develop an effective agent it may be co-opted over time by the virus changing its genetic matrix to cope with the agent of destruction.
Tops down refers to eschewing the isolated chemical specificity approach by trying to find a way to enhance the intrinsic disease fighting capability within each individual organism in contrast to fighting disease across the collective similarities in all human genetic structure. We have an immune system mechanism which is perfectly capable of handling almost all, if not certainly all, known diseases. Variations in genetic structure from human to human make humans genetically more or less susceptible to viral machines and our immune systems varyingly capable in handling an invasion or expansion. If we can engineer agents which enhance a given defective genetic structure by adding functional components to the cellular entities and their assistants whose job is to destroy the invading virus, then disease which the given individual's immune system can't handle, with component enhancement would be able to handle.
In both cases we're engineering drugs, chemicals, which have a certain effect. Most of the time even with significantly strong diseases, the best action people can take is to let their immune systems handle it without encumbrance. You can't realistically expect to improve on the efficiency of our immune system's response, so it is folly to take drugs in most cases when you get sick. You just compromise what nature has spent several billions of years evolving and which fight or manages millions of micro organisms entering your body daily. Maybe in 10,000 years when we have begun to scratch the surface of the complexity of genetic machinery, this will no longer be the case. For the forseeable future we can only supersede the immune system's functionality when it has a genetic deficiency precluding an adequate response to invading viral machine. The circumstance is life-threatening so we are justified in making such attempts.
The tops down approach doesn't try to make too many unwarranted assumptions about how things work at the chemical cellular level. It tries to build on what is known more so than the bottoms up. The bottoms up is more of a "Russian roulette" approach, because after you've reached a probable good drug either by intellectual cogitation or by random searching guided by considerations of known pathology and chemical morphology, you still have to test it. There is a better chance for a drug to succeed which is built on your genetics than one that is built on everyone's shared genetic similarities. The almost infinite possibilities of rational construction make the bottoms up approach very narrow in its applicability and practicality.
IMNR is the premier play in tops down. Hope that clarifies the issue a bit, but obviously the notions overlap and are vague in their definitions. |
