Cimora
Bluelighter
- Joined
- Mar 20, 2000
- Messages
- 2,495
I drove back and forth with a guy I knew who was doing neurobio research at the Scripps Insitute there. He was working on a simulating just the electrical properties of a single neuron. Just one; it took a while to simulate only a few seconds of activity, and God knows how accurate it was. In the human brain there are 100 billion neurons -- you would need hundreds of gigabytes of memory merely to list them all....
This is a great example of the kind of level of detail a functional model really needs. A single nueron, with all its charges, chemicals and gates is quite complex indeed. But the function it serves - to activate at a set level depending on the strenght of an association can be emulated perfectly using neural networks. The have built fairly good learning AI on the back of this simple multi-CPU model - an insect robot which can learn to navigate its enviroment and how to walk from stratch has been made, they are going to use it for a mars mission. So, in pratice the emulation doesnt need all that detail to work in the same way. Biology is based on legacy and branching technolgy, often its pretty ineffeciant compared to how wed have designed the same thing.
On the heart model -- not sure what you're talking about, but there are many heart models out there, some better than others, all very limited and incredible simplifications.
Ive posted some links at the end, that may help. Check em out, theres actually alot of links out there on the virtual human heart.
On the contrary, they're pretty much the only way we can learn about biology. Look, pick any recent advance in large-scale (eg more than just pure biochem/molecular bio) biology, and try and tell me how it could have been achieved without the use of animals models. Can you?
With that statement i was refering to purely brain studies. There are some low-level things we have learnt from animal brain studies, but most of the big stuff is human only learnt. That stuff we have learnt from animals have involved a very high level of suffering for the animal subject, making some of these low-level findings questionable in terms of ethical input and output in my mind.
But take mood. Mood in humans works not much like any other animal - mood is a composite of primitive emotions composed by the frontal cortex. Langauge - theres little we can learn for animals. Conceptualisation, logic, anticipation, high level perceptual processes. This list goes on, and these are some of the things we really want to know about ourselves.
You wanted an example? memory. All recent advances in understanding concious memory in humans have come directly from human case studies. Many regions of neuropsych are like this: theres only one way to really learn, humans; damaged case studies and brain imaging.
But in many cases even these drugs aren't on functions specific to humans; usually they target older systems and mechanisms in the brain that are fairly conserved across all mammals. I don't know of a single drug that affects the kind of high-level cortex function unique to humans (probably with good reason -- no animals to do research on!) Most mammals have an anxiety system built in the amygdala; they can all have seizures, they all sleep, etc
There is some truth in what you say, all higher functions are based on and utilise more primitive functions. We are a legacy machine. But take any of those functions i mentioned are there is some profound variation from that base. Mood in humans is more complex, mediated through the frontal cortex - trying to emulate human mood with another animal is like trying to cook a roast with a toaster. For epilepsy, the seizure is local to a region of the brain, thus the way it spreads and the functions in effects are unique to the specifics of a brain - studying other humans may even be useless to some degree let alone animals. Admited sleep and anxiety are fairly basic though. But you see, alot of these drugs are being tested for safety on primates and mice - when many of them such as anti-depressants, anti-pscyhotics are basically human specific, the animals tests are more or less a waste of time and life. (Which i find particularly distasteful considering primates are often used. They are our ancestors for christs sakes, nearly on the same level of evolution and 99% genetically similar - they experience psychological pain - we should not call apon them for animal studies unless the circumstances truely justify it). You want a medicine that works on higher functions, actually most anti-psychotic and anti-depressant medicines do. And yes, they are tested on animals, but god only knows why.
Research is very rarely done with monkeys, partially because of cost, partially because of ethical concerns. In safety trials, typically rats & rabbits would be used, or mice & pigs.
In most science yes, your right. And rats and bunnies dont have true psychological pain, so thats more acceptable to my mind. But monkeys are used alot in brain and neurological drug research - IMO needlessly. Which animal you use depends on which is "the most compatible". The problem is, is that the most compatible isnt always very compatible at all.
You and vegan both seem to think that because there's not a perfect correlation between toxic effects in animals and people, there must be no correlation at all. But there'a big range between 0% and 100%! Some quick googling reveals that 70%+ of toxic reactions are caught in animal trials.
I think that would be a hard thing to study with objectivity (how much is actually caught). The FDA certainly wouldnt want you to have to low a number

Im not saying less than perfect is unacceptable. 99% safety is good safety IMO. Its just that, in practice there are many specific areas we cant give reliable safety for - and people assume we can, so they waste life, time and money on animal studies they dont need. Im with you on using rats to cure cancer - I just think we should be prepared to say sometimes - look, we cant know how safe it is, until we give it to someone, otherwise we are fooling ourselves.
And if we are talking about ethics, it should balance something like this:
Benefit of animal testing (ie the suffering eased) TIMES the accuracy SHOULD BE LESS THAN the suffering of the animal.
If people arent doing an equation everytime they sacrific a life or cause pain to acheive a greater good, they have lost sight of their goal. Thats why the exact degree of safety is important IMO - its part of the cost/benefit analysis.
Can you point me to your reference for this? I saw a few papers suggesting something along those lines a couple years ago, but they were all extremely sketchy, borderline effects. I thought they'd all failed to find any support for the idea -- has there been some new research I missed?
It was about a year ago, on this site i first saw that research. It was replicated once, and they studied multiple subjects for the axon damage (of course to find actual axon damage they needed dead long-term prozac users so this research is only just becoming possible, and the subject count should be a bit higher before we are certain about this). It may not be conclusive, but one or two replications more and we may have to accept this as fact.
As far as i know, no one has ever studied long-term prozac users and NOT found axon damage post-mortum - thats what would disprove this.
It makes sense though, that excesses of the most important neurotransmitter in any form could be trouble.
There are plenty of other examples of things declared 100% safe only to be deadly or damaging. Thats part of the problem: the public perception that safety testing is 100% reliable.
And ive saved the best till last:
Firstly, heres a good use of animal testing to acheive simulation, here they are using slices of rats brains to create an artificial hippocampus (the world first brain prothesis) - something that will easy suffering and further understanding: http://www.newscientist.com/news/news.jsp?id=ns99993488
Secondly, heres a project to virtualise the human body, which mentions the heart simulation ive meantioned:
http://news.bbc.co.uk/2/hi/health/345762.stm
Virutalising the human body just isnt possible yet? maybe you should tell these guys

Last edited: