In my recent rehashing, rebranding, and repurposing an article addressing many of the flaws in the so-called scientific arguments against animal research often made by animal rights activists and extremists, I only briefly discussed one common argument among many, namely that computer simulations can replace the use of animals in medical research by modeling human physiology. I pointed out that, for a simulation to be valid, we have to understand quite a bit about the system beforehand and that simulations aren’t much good if they can’t be tested against reality. Mark Chu-Carroll decided to weigh in, and he did it in much more detail than I could. He also did it in a lot more detail than I could, because, well, he’s a computer geek and computer simulations are one thing he can do:
Let’s start at the beginning: just what is a simulation? It’s a model of a real system, which attempts to reproduce the effects and/or behavior of the system that it models. In the case of computer simulations, which is really what we’re talking about, we produce a mathematical model and algorithmically describe how the model evolves over time. In other words, we write a computer program that runs a mathematical model of a process.
And right there is the problem. We produce the model, and we implement the model. It does exactly what we tell it to. We programmers have a saying which applies particularly well to simulations: garbage in, garbage out. Computers do exactly what you tell them to; if what you tell them to do isn’t right, then no amount of computer power is going to change the fact that you didn’t tell them to do the right thing.
Precisely. Mark then goes on to explain why.
Add computer simulations to the list of “alternatives” touted by animal rights activists and/or extremists whose efficacy they never manage to show to be greater than the methods they’re supposed to replace.
67 replies on “Computer simulations to “replace” animal research: Garbage in, garbage out”
Eventually we will have computer models to do this sort of thing with, in all likely hood. I wonder how many people will be harassed and stalked and threatened and terrorized for doing the necessary animal research that goes into making a model? It’s like the stem cell deniers, who stamp and shout about how only adult stem cells are being used for possible treatments, ignoring the fact that we only knew what to do with those adult stem cells because we had studied embryonic stem cells.
Sad day.
Orac,
The legal case against animal experimentation is already well-established, as you know. All that is left is the balancing test of whether the pain and suffering caused to the subject animal is overridden by the value of the information which might be gleaned from its abuse.
This post is not helpful to your cause, since it basically says we’re not smart enough to develop research techniques that do not require lethal and invasive experimentation, or to be less generous, we haven’t even really tried, nor do we intend to.
Douglas,
Your last statement is insulting to anyone who works in biological simulations. But they would be the first to admit the limitations and problems that simulations have. You have no reason to think that, nothing Orac says here indicates that is the mindset here or what is going on. You also address none of the points actually made, the fact that animal research is needed to make these simultations, etc. A big part of me wonders if you actually understand what was said.
Douglas Watts: By that same token, we’ll never be “smart enough” to break the speed of light.
Is moving the goalposts and ignoring the evidence your usual response when Orac spanks you?
Really?
That’s news to me. And here I thought my research was legal, having been approved and funded by the NIH and approved by our local IACUC and all.
Douglas,
Way to completely miss the point. Testing such simulations to determine whether they’re accurate enough for a particular purpose requires comparing their results to reality. IOW, they are valuable only if and insofar they reproduce the results obtained from real (i.e. animal) experiments. Determining whether they reproduce the results of an experiment is logically impossible without performing said experiment first.
There’s no element of “we’re not smart enough” involved. You’re essentially demanding that we must have complete, 100% accurate, understanding of all science as a starting point for science.
Orac @#5: Wow, so Douglas Watts is out-and-out lying? Color me shocked. Not.
Douglas,
Feel free to help us dummies out, won’t you? If you or someone you know has sufficient biological information to program a computer to simulate a complicated multicellular organism, bring it on! Don’t let the fact that we professional biologists aren’t ‘really trying’ deter you from coming up with your own alternative. Post your results here so that we researchers will know how to access your wondrous new invention! Oh, and please be sure to carefully describe the methods you use to a) glean sufficient information to effectively program the computer and b) verify that your artificial system is behaving as a biological system would.
I eagerly await your breakthrough!
If the activists think they can write the computer simulation, they should do so.
And tell us when they are embarking on their goal. It will be fun to watch people who’ve never heard of the halting problem try to write an algorithm they don’t even understand.
Douglas doesn’t even need a computer simulation. He already knows EVERYTHING about every single biological system.
We can all quit our jobs because bioscience has now been completed.
Douglas,
To be a little clearer, and slightly less insulting (not that you don’t deserve it) than the other commenters:
to accurately model the human body such that the model would accurately predict unforseen complications, you would literally need to model every atom in the human body as well as every atom of substances commonly entering and exiting the body as well as common imperfections in human biological systems. You would not be able to simplify anything, or make any assumptions or you would miss potential complications. That’s the thing about unforseen complications…. they’re unforseen, hence we cannot easily model them.
Why did physicists build the Large Hadron Collider anyway? Isn’t all you need a computer simulation of the “stuff you don’t know in the first place”?
Too much Star Trek…
animal research is vital for effective, safe, accurate research in biomedicine and biology in many ways.
Specifically because different animals are like us in different ways, we can sometimes use mice, sometimes pigs, sometimes primates.
your wacky poetic website http://tispaquin.blogspot.com/
has a disclaimer in the middle of arguments being ‘refuted’ in the name of animal rights.
_____________________
“5. Animals are not like us, therefore lethal animal research is okay if it may, perhaps, benefit humans.
This could justify anything done to animals, even the most extreme cruelty. It is the old “they are not like us” argument which has been used to justify endless amounts of cruelty by humans against other humans who were also deemed “not like us.”
________________________
You (or someone, perhaps Ms. TISPAQUIN?) miss several points, CAMILLE….
Animal research is effective BECAUSE the animals are much like us. However, in an important difference, they are NOT sentient creatures.
Research animals are indeed almost always, in regulated modern countries, kept in highly regulated, very pristine and humane conditions. Better than pets, better, in too many cases, than many children or elders.
But to somehow link the use of God’s gift to us, in the way of animals that biologically mimic many of our biologic processes, to the historical racist/xenophobic dehumanizing of “enemies” to justify discrimination, killing, etc IS A FAR FROM ACCURATE ANALOGY… and a bad example of poor reason.
Here is what Doug said directly to Orac in another blog:
I’ve use computer modeling (simulations) quite a bit in microbial ecology and geomicrobiology. The problem with modeling/simulations is that you have to know everything about the system in order for the model to accurately predict what will happen in novel situations.
The more you know about the system (and can include in the computer simulation), the more accurate the model will be. Conversely, if you don’t include (or don’t know about) some components or interactions in the system, the model will give you inaccurate (and sometimes nonsensical) results.
In my use of modeling, we would use the model to make predictions about what would happen, then we would perform experiments to see how well the model compared to “real life”. We would then use the experimental data to “optimise” (tweak) the model. Even when we had the models optimised, they still didn’t predict terribly well. They were good to within an order of magnitude (factor of ten) – generally. For our purposes, this was enough to allow us to design future experiments, but I wouldn’t have used any of the model data to make predictions that lives would depend on.
These were extremely simple systems using relatively simple organisms (bacteria) whose physiology we knew in extreme detail. We don’t know too many multicellular organisms that well and we certainly can’t predict how humans (who happen to be much more genetically diverse than lab animals) will respond.
Once we know the developmental physiology, metabolism and genetic interactions of humans to the degree we now know those of, say C. elegans (a worm), we might be able to construct a computer model. Bear in mind, even the folks who study C. elegans haven’t been able to develop a computer simulation that will predict how the worm will react in all potential situations.
Mr. Watts seems to have confused cannot with will not. We cannot (yet) construct a computer simulation of human physiology – it’s not a question of will. Two major hurdles stand in the way:
[1] We do not yet have a comprehensive understanding of human physiology (or even of rat physiology), so we lack the knowledge base needed to programme a simulation.
[2] The number of potential interactions in a complex organism (like humans or rat or even simple worms) is so large that even the most advanced supercomputers of our time could not carry out such a simulation (presuming we were “gifted” with the necessary information to programme it).
Again, it is not a lack of will that prevents us from using computer simulation as a substitute for animal experimentation. What is lacking is the knowledge and the technology.
To get the knowledge, we need to do more animal experimentation.
Prometheus
Prometheus has it right. I’m coming at the problem from the Computer Science side, but have he same conclusions. The reason that we need to do so many animal tests is that life is really, really complex.
If the pharmaceutical companies had enough knowledge of the mechanisms of life, drug discovery wouldn’t be such a gamble. Most drugs that seem promising in the lab never make it to human testing, because of unforeseen (and currently unforeseeable) interactions.
This brings to mind the famous quote by George Box: “Essentially, all models are wrong, but some are useful.”
*laughs*
We can’t even create a computer simulation to accurately model reentry heating on a Space Shuttle, an entirely engineered object, and people seriously think we can accurately model the vastly more complicated chemical and mechanical workings of the human body, which doesn’t even come with a high-level technical drawing, much less a user’s manual or detailed system specifications?
I agree with Koray @ 12 — people have been watching too much Star Trek if they think we can model everything accurately enough to do away with animal research. Seriously, I think the only alternatives to animal research are to either stop making medical progress, or start testing on humans. And frankly, if it’s gotta be us or the animals, I know which one I’m picking.
I used to work in particle physics. In that field we make extensive use of Monte Carlo and detector simulations in order to try to understand the background and signal before an experiment is done, and so that we can make an attempt to extract signal from all the noise when it is running. We know a lot about this and the models are pretty damn good but not perfect, there are lots of problems with them. While we have simulations of the detector components people still go out and run test beams to check how they actually perform in real life. Normally we have a good idea about the general behavior but there are always lots of weird things happening in real life once you plop that calorimeter segment in front of a beam, real life is complicated and full of aberrations.
As someone who does lots of math and computer modeling of biological systems, I can tell you that they will almost certainly never replace the need for real data from real systems (e.g. animals, people, plants, and so forth). Can they help reduce that need? Yes. Point out places where we need more data from real organisms? Certainly. But replace experimental work permanently? Never going to happen.
Mathematical or computer models aren’t like magic eight balls. They don’t miraculously predict the future. They are a precisely formulated mimics of how we think a particular system works, often constructed to answer specific questions, and usually pretty bad at answering questions for which they weren’t constructed.
Whatever knowledge one squeezes from a computer or mathematical model is, has been, and always will be rooted in solid empirical evidence.
“It is the old ‘they are not like us’ argument which has been used to justify endless amounts of cruelty by humans against other humans who were also deemed ‘not like us.'”
Yes, because the lives of the Jewish people murdered by the Nazis have the same value and importance as those of rats.
Oh wait, that’s what the Nazi’s thought…
Paul, thanks for your post. Simulations are based on what we think is happening. If we understand the system well we can often simulate the behavior and actually see some interesting things that go beyond the work that was used to make it. You can see some unexpected things but you would probably not run out and act like you discovered something if you saw this, you would have to actually do experiments to confirm what your simulation was saying.
I’m bothered by the implicit ethical problem posed by Watts et al.
If a computer simulation is a sufficiently faithful model of a human being to (for instance) predict cognitive disorders, learning disorders, etc. — necessary to enable us to use it for things like drug testing — then hasn’t it satisfied the Turing Test and qualified itself as at least as “human” as the animals we’re using now?
It seems to me that by the time we have computer models useful for biological research, it will be unethical to experiment on them.
Okay. Let’s start from the beginning. The abuse of animals is prima facie unlawful. A narrow exemption in animal abuse law has been created for licensed animal research, subject to approval of a lawful permit and full compliance with all conditions in the permit. Like any license for an otherwise unlawful activity (think driving on a public road), the license conveys a privilege, not a right. There is no right to abuse or experiment on animals.
The entire purpose and intent of laws prohibiting animal abuse and maltreatment is to stop the abuse and maltreatment of animals. The exemption granted to licensed animal researchers is in recognition that, in exchange for the promise of certain eventual societal benefits, and subject to strict controls, animals can be treated by researchers in ways that would never be allowed under the law in any other circumstances.
The exemption for research is couched by an understanding that the researcher has an affirmative obligation to seek and use the least invasive and most humane methods available. If the researcher does not wish to play by these rules, they are always free to not use animals in their research and do something else. So the arc is always towards the use of the most humane and least invasive techniques available. The laws imply constant improvement, not stasis, inertia and stagnation. Investigation and proposal of new and more humane techniques by researchers is always encouraged. The onus and burden of proof is always on the researchers, who are conducting this research voluntarily. This is the legislative schematic.
The cornerstone of the very laws which allow any animal experimentation to occur focusses on the welfare of the subject animal, which certainly does not want to be in the situation it has been forcibly placed in.
These laws are not about the welfare of the humans who might, 10 or 20 years later, possibly benefit from the information gained from lethal research on animals. For good reason. The effect on the subject animals is acute, irreversible and knowable. The intended benefit for humans will not be discerned for many years or perhaps never. One outcome is a certainty (the animal suffers and dies). The other is a possibility (some human may or may not benefit). It is for this reason animal abuse laws are rooted in the concern for the subject animal.
If the animal research community is in agreement that no improvements beyond the status quo are technically feasible, then they need to say this up front now, as in “We Have Failed,” because researchers have always promised the public and legislators that new and improved techniques are just a few years away that would eliminate the most egregious types of lethal animal research. If this is not the case, if the problem is too hard to solve, or if researchers have just plain given up trying, that information would be useful to the public and legislators so we can reassess the underlying laws which allow any animal experimentation to exist.
How much animal testing will be required to create a reliable simulation? Simulations have to be checked against the reality of what they simulate.
How much animal testing will be required to validate and test simulations as they are altered? Any alteration will require testing to make sure the changes remain accurate to actual animal biological responses.
On the other hand most animal testing is done as a substitution for human testing. More human testing equates to less animal testing. If the animal rights protesters would just volunteer for human testing they could make a significant dent in the numbers of animals used.
Animal rights groups might look into lowering the barriers to testing on humans. The free market folks would likely go along. With about seven billion people in the world the value of any one human is going down. Continued to the logical end humans will be the cheap and effective substitute for more expensive animals.
At least until we have a reliable computer simulation for a human. At which point the computer simulations will be testing on humans to spare simulations the pain and suffering. No doubt some of those simulations will be protesting testing on humans and …
I wonder if Mr. Watts would fly on an airplane that had only been tested through computer simulation. There is a greater understanding of structural analysis than biology, but still new structures are always tested.
Douglas:
At this moment, I’m inconveniencing and de-homing a lot of animals by sitting in a comfortable habitat I won’t allow them to enter, thereby negatively affecting the quality of their lives. I’m aware of it. That still doesn’t mean I’m going to leave the front door, windows, and fridge open, or that I will allow it when mice home themselves in my walls (I rehome them with a live trap, but I’m sure this causes them considerable distress and discomfort.)
I’m not sure that animal abuse laws are rooted in concern for the subject animal itself as they are in concern that humans not be inhumane.
No improvements past a lot of status quos are currently feasible. I don’t have a flying car; the foster-child system has a lot of problems. That doesn’t mean that you throw out everything; it means you go on with what you have. A problem being too hard to solve immediately doesn’t mean that we don’t have more important problems that, unfortunately, we need everything in the toolbox for.
Mr. Watts:
Please provide a citation for this claim, along with the legislative definition of ‘the most egregious types of lethal animal research’. Thanks in advance.
You can write simulations to find out certain aspects of the way things work that you didn’t already know. In climate science, for example, they do this all the time.
With living organisms, we’re talking about and entirely different level of complexity, however. And you can’t just apply the laws of physics to living organisms and hope to get a result. I don’t think they thought it through very well.
Douglas Watts wrote:
“The abuse of animals is prima facie unlawful.”
False. There are no laws prohibiting the abuse of virtually all animals. Laws only protect the animals that we have decided are more valuable than others.
“A narrow exemption in animal abuse law has been created for licensed animal research, subject to approval of a lawful permit and full compliance with all conditions in the permit.”
False. In the US, there are no such permits.
“Like any license for an otherwise unlawful activity (think driving on a public road), the license conveys a privilege, not a right.”
There are no such licenses in the US. You are one clueless puppy.
“There is no right to abuse or experiment on animals.”
Rights aren’t relevant. Why do you use the particularly deceptive phrasing “experiment on animals,” Doug? Why does the animal “rights” movement ignore, and even aggressively lie to conceal, animal suffering and death in experiments in which they aren’t used as subjects, but as tools?
Why is it all about experiments, but the animals exploited (as mere tools) to produce clinical assays on human subjects are utterly ignored? Don’t you believe that they have rights, too?
“The entire purpose and intent of laws prohibiting animal abuse and maltreatment is to stop the abuse and maltreatment of animals.”
Go on, Captain Obvious…
“The exemption granted to licensed animal researchers…”
Doug, I’m one of those “animal researchers,” but I’ve never had a license. You’re just making this up.
“… is in recognition that, in exchange for the promise of certain eventual societal benefits, and subject to strict controls, animals can be treated by researchers in ways that would never be allowed under the law in any other circumstances.”
This is a ludicrously false claim.
I can abuse mice at will in my home. I can go to a hardware store and buy an incredibly cruel glue trap, catch mice with it, and leave them to starve. I could never get IACUC approval to do anything like that in the lab.
“The exemption for research is couched by an understanding that the researcher has an affirmative obligation to seek and use the least invasive and most humane methods available.”
Yes, we do. So why does everyone in your movement claim that cell culture is a “non-animal alternative” when it is neither?
Doug Opined “The abuse of animals is prima facie unlawful.”
Apart from the misuse of legal jargon here its a good thing that this is bullshit otherwise farming would be out of the question.
‘researchers have always promised the public and legislators that new and improved techniques are just a few years away that would eliminate the most egregious types of lethal animal research.’ Please provide a citation for this claim, along with the legislative definition of ‘the most egregious types of lethal animal research’. Thanks in advance.
—
You’re welcome, Jen. The promise of less invasive and more humane research is implicit in the statutes which allow animal research to occur. The implicit assumption is that scientists do not want to harm and hurt animals under their care and are always seeking any way possible to minimize or eliminate any pain and suffering their research might cause. As the sole caregivers and stewards for these animals’ lives, this is the researchers’ prime responsibility: to do no harm.
By definition, lethal animal research is egregious. A research technique which renders the subject animal dead or in such pain and suffering that it needs to be euthanized is by definition the harshest treatment you can render to an animal. A research technique which allows the animal to recover and be re-acquainted to the world of non-laboratory test subjects and allowed to live out its life in its natural habitat is obviously the most preferable outcome. In this category I would include surgically removing an anal fin from a juvenile Atlantic salmon parr for the purpose of a micro-satellite genetic assay. The intervention is elective, has some, but minor, effects on the animal, and is posited upon gathering knowledge which ultimately will help protect the subject animal, its siblings and species.
My professional experience in this field is protecting species in danger of extinction, particularly Atlantic salmon, shortnosed sturgeon and Atlantic sturgeon. Unfortunately, chimpanzees and bonobos and apes now fall into this category. As you are probably aware, if chimps, bonobos and great apes were native to the U.S., the Endangered Species Act would fundamentally prohibit any elective experimentation on them for any purpose except to keep them from going extinct.
@Douglas Watts: please cite the law by statute number.
Or even cite some case law on the topic.
I’ll help you: federal laws can be found here :
http://www.law.cornell.edu/uscode/
“The abuse of animals is prima facie unlawful.”
Apart from the misuse of legal jargon here its a good thing that this is bullshit otherwise farming would be out of the question.
—
Not true. State laws are quite specific on husbandry standards for all farm animals. In Maine, where I live, animal abuse laws cover farm animals and set very specific standards for their care and welfare, much in the same way as pets. If farm animals are not cared for properly and humanely and within legal guidelines, the farmer can and will be charged with animal abuse, which carries jail time.
In some states affected by lead poisoning and/or factory farming lobbyists, the legal standards are probably much more lax than in Maine. Maine is a very rural state but animal abuse is not tolerated here. Most farmers here are very proud of how they care for their animals. It’s an ethic.
@22:
That’s if, and only if, we try to simulate brain function completely. Which, as you say, would be the only way to do it if you wanted to study behaviour rather than the more mucky internals. Fortunately, in most situations you can treat the brain as a black box that issues electro-chemical commands as well as various other chemicals.
If we were to produce a simulation of the human body that solely concentrated on drug interactions, building a full neural net wouldn’t be strictly necessary, only nice to have (if you ignore the ethics). Of course, making even a simple drug-interaction model is well beyond our capabilities…
umm, isn’t this kind of ironic- your extolling the importance of animal research yet arguing that monkey studies are somehow not valid to be doing in terms of vaccine safety (i.e. Hewitson’s research)??
Douglas
Farm animals are killed and eaten and their body parts are used for things like shoes. American farming practices which include pen feeding are regarded as ‘abuse’ by many.
Crop farming also often involves ploughing which if you’re a worm or a vole or a field mouse is going to cause you pain, suffering and death.
Hunting and fishing would be out of the question to.
You keep making stuff up and then moving onto making more stuff up when the previous instance has been pointed out.
No. He’s extolling the importance of animal research done correctly and for the right reasons. Hewitson’s study was so horrible that anyone with a basic understanding of how studies are created can tell they screwed the study up. The controls were unbalanced. The controls actually changed.
Doug’s lie:
“The abuse of animals is prima facie unlawful.”
Antipodean’s challenge:
“Apart from the misuse of legal jargon here its a good thing that this is bullshit otherwise farming would be out of the question.”
Doug moves the goalposts:
“Not true. State laws are quite specific on husbandry standards for all farm animals.”
You didn’t specify “farm animals,” Doug. What are the laws prohibiting the abuse of, say, rats by poisoning them with warfarin just for fun?
“In Maine, where I live, animal abuse laws cover farm animals…”
There are laws to cover pets (personal property), and there are laws to cover game animals (collective property). There are no laws at all relating to the abuse of the vast majority of animals.
“… and set very specific standards for their care and welfare, much in the same way as pets. If farm animals are not cared for properly and humanely and within legal guidelines, the farmer can and will be charged with animal abuse, which carries jail time.”
The same farmer can shoot, poison, and torture pest animals without a care in the world.
“Most farmers here are very proud of how they care for their animals. It’s an ethic.”
As are most researchers. It’s an ethic. Why don’t you understand this, Doug?
“In Maine, where I live…”
…Fffffffuck.
Anyway, I notice that Douglas Watts has performed a successful threadjack, once again proving that animal “rights” activists can’t and won’t defend their “scientific” claims.
As for computer simulations, the ARA’s are ignoring one crucial detail: simulations have to be checked against empirical data to gauge their robustness. How do they propose we obtain this empirical data? You’re essentially back to square one…
Heck, I’m an electrical engineer in training, and even simulating fairly simple electrical circuits is a pretty error-prone process. It’s pretty easy to construct a situation in which SPICE or its descendants will throw up its hands and tell you (in cryptic error messages, but still) that its models/algorithms could not produce a result for your circuit. The human body (or the mouse body, for that matter) is presumably far more complex than anything SPICE has ever been expected to deal with.
On that note, I’d point out that “garbage in, garbage out” applies just as much to the combination of good models and bad/inappropriate data as it does to bad models and good data (a concept that seems especially elusive to some global warming deniers). Complex systems are virtually always modeled with simplified assumptions that hold under certain conditions (which is why we still teach and use Newtonian mechanics instead of jumping straight to relativistic mechanics, even though Newtonian mechanics is wrong). Knowing those conditions and being able to figure out whether they hold for a given set of inputs to a given model is often the deciding factor in whether simulation results can be trusted. It can take centuries of experimentation to discover how the models fail, even for a question as seemingly simple as “how fast does an object go if you push on it with a given force?”
I could write a comment here about how the anthropogenic global warming enthusiasts play fast and loose with models of a very complex chaotic system, but I won’t, not wishing to shake the faith of their parishioners and have them get all mean and nasty.
Please don’t fall for the obvious flamebait. That is all.
Aw shucks, the experiments of the Nazis and Japanese on human subjects should be seen as the least worse form of suffering during the war. At least they were engaging in the furtherment of medical science – others were just killing and torturing for the fun of it.
Tyler, sorry, that’s what I’m here for.
Justthismoron, go suck a dog’s ass.
@justthisguy
Well, climate models are at the very edge of what is possible if you have thousands of scientists, the best and fastest computers of the planet, data from centuries and from all over different aspects of the system. And if you make simplifications. And if your system is fairly closed. And you have just one system.
Its possible, but barely so.
But organisms are vastly more complex, so whats your point?
In general, it seems to me that scientists are much more aware of the complexity of living systems, i.e. animals, than so-called animal rights activists. If they want to replace testing with models it just shows that these activists have no idea what an animal is. They just see it as a means to fulfill their own agenda. How else can you explain such a completely out-of-this-world stupidity?
bcpmoon:
And they still use FORTRAN, due to lots of legacy software (and it does do well with large scale scientific computing, especially vector computing).
> We produce the model, and we implement the
> model. It does exactly what we tell it to.
> We programmers have a saying which applies
> particularly well to simulations: garbage in,
> garbage out. Computers do exactly what you
> tell them to;
I disagree.
With evolutionary programming / genetic algorithms you can make the computer do what you want without telling it what to do.
For example someone wrote a pretty simple program to evolve simple creatures. And it works.
the frog
or the end-over-end-worm
Garbage in -> something great out.
Douglas has made the following statement “if the animal research community is in agreement that no improvements beyond the status quo are technically feasible, then they need to say this up front now, as in “We Have Failed,” because researchers have always promised the public and legislators that new and improved techniques are just a few years away that would eliminate the most egregious types of lethal animal research.”
This is pure nonsense, the assumption that scientists must be able to say for sure that a new technology will be able to replace particular animal procedures for animal research to be ethical is asking the impossible. Even for many areas wher scientists are fairly sure that development of non-animal alternatives is possible, for example batch testing of botulinum toxin, the time scale within which this will happen is not certain, might be 5 years, might be 10. That doesn’t mean thaat efforts are not being made to develop replacements, indeed by looking through the links below you can see that some of these efforts have already been successful or are expected to be in the near future.
http://www.nc3rs.org.uk/
http://altweb.jhsph.edu/
http://ecvam.jrc.ec.europa.eu/
http://iccvam.niehs.nih.gov/home.htm
Of course these are only the big players, and are as much if not more concerned with co-ordinating efforts of industry and academia as with directly funding their own research.
Douglas Watts is a classic example of clothing unreasonable demands in very reasonable sounding language, something that those of us familiar with anti-vaccination campaigns and Intelligent Design creationism are all too familiar with.
Jen: The Hewittson monkey study isn’t useless because it used monkeys, but because it was a poorly done study with a low sample size, too few control animals, questionable blinding and it was run by people who have an undisclosed conflict of interest. Orac has written extensively on this. I know it’s hard, Jen, but please try to keep up. Ever thought of a reading for comprehension class ? It’s okay to make notes, you know.
Douglas Watts: Still waiting for those law cites, how’s that coming ? Statutes will do, if you haven’t time for case law as well.
Decades ago, during development of a better “artificial skin” for treating deep burns, the researchers used computer simulations.
They knew from previous experiments (on live animals, including humans, and in tissue cultures) what each layer’s physical characteristics should be to best replace skin and to let new tissue grow into the matrix of material. They knew from previous experiments and data from the manufacturers (test tube stuff) about the physical characteristics of thousands of potential membrane materials.
Running an extremely long datamining episode (something my current computer could do in a few seconds took days back then) brought a couple dozen candidates to the top of the heap for a closer look. Many more computer runs tested the possible combinations and thicknesses of the layers.
It saved a tremendous amount of time and expense by reducing the number of materials to test on the pigs, and hence reduced the number of pigs. But it did not eliminate the need to test the materials in living organisms to make sure the predicted behavior was what actually happened. BTW, the “best” candidate after the computer simulations was not the best overall solution in live animal tests. I don’t remember what the problem with it was, but one of the other combos had the best overall results when they balanced speed of new skin growth and quality of skin. That is something that could not be “modeled” then because there was no data to use in a model.
Using lab animals is/was NOT cheap, even back then under the fairly light USDA restrictions. These were long-term experiments to see how well skin regrew over a period of months, so these were among the best-cared for piggies in the USA.
Using large, intelligent animals such as pigs in research means you have to persuade them to cooperate, and they remember who treats them and how they are treated. Maltreating animals is pointless, if only because it makes the research harder to do. I had less trouble getting blood samples from the pigs than I did from toddlers in the peds wards because we trained the pigs to tolerate it by using treats and attention.
Even after the release of the artificial skin, research continues: new materials, more clinical data (from burn patients and lab animals), more basic research about skin … the computer models are many generations beyond what I worked on, the predictions about materials are better. The artificial skin is also far better.
PS: It’s being used in animals by vets, to treat burns in domestic and wild animals.
DLC “The Hewittson monkey study isn’t useless because it used monkeys, but because it was a poorly done study with a low sample size, too few control animals, questionable blinding and it was run by people who have an undisclosed conflict of interest. Orac has written extensively on this.”
Actually as Orac and Prometheus pointed out the problems with the control groups went well beyond small size and lack of blinding, there were indications that the controls were not even treated at the same time as the experimental animals.
http://photoninthedarkness.com/?p=178
As I said in a comment to one of Orac’s previous posts Wakefield and Hewitson may not have tortured the monkeys but they certainly tortured the data. That’s why the study was unethical. There’s still a question mark hanging over whether the study, as it was actually conducted, had IACUC approval…Hewitson appears to have departed Pitt shortly after conducting it under something of a cloud.
Am I the only person to notice that Douglas Watts just kind of shifted his commentary when he got spanked for bullshit and then disappeared when it became apparent that no one was going to accept his bullshit without evidence? Douglas’ debate style is very reminiscent of the style I recall other kids and I using back in elementary and part way into middle school. Only somewhere in seventh grade we started to learn about persuasive writing and at the very least had an idea that this style Douglas is using was the wrong way to do it…
@28 You can write simulations to find out certain aspects of the way things work that you didn’t already know. In climate science, for example, they do this all the time.
The simulations for climate science are checked against real-world climate behavior, then the simulation software is corrected for the new data and run again … lather, rinse, repeat.
Would you want the drugs you use on yourself and your children to be tested by models that have the same accuracy as the current hurricane tracking systems? The current tornado prediction systems? The 10-day weather forecast?
from a much longer, rambling entry on douglas’ blog “refuting” pro-research arguments:
“4. Much 20th century medical knowledge is based on animal research.
Because so much lethal animal research has been conducted in the 20th century, it is inevitable that a lot of medical knowledge is partly or wholly drawn from this research. A statement of fact is not an argument.”
A statement of fact is not an argument.
I guess that explains the lack of citations. I’m a mainer, I grew up next to a farm. I am against unnecessary animal cruelty. The level to which this dude takes it is ridiculous.
@DuWayne
You’re not the only one who noticed. I think we actually won the game of whack-a-mole!
Ah, see now, if you’d been smart like our good buddy Andrew Wakefield, you would have also trained the toddlers to give you blood samples by offering them treats and money.
The point about animal research ultimately benefiting animals by advancing the state of veterinary medicine is useful. I feel I have to mention the bags of dogfood I have occasionally seen which proudly state “not tested on animals”.
Oh, so you don’t even know if they *like* the food? Jeez…
😛
Um, isn’t it kind of ironic that you are using the English language to extol your stupid ideas, yet arguing that the words Orac says in English are somehow not valid?
Um, isn’t it kind of ironic that you are using a computer to post comments to a blog, yet arguing that some other people’s blog comments are wrong?
Um, isn’t it kind of ironic that I am perfectly okay with eating apples, but arguing that eating arsenic is somehow not a good idea? After all, both propositions involve eating things beginning with the letter A!
Um, isn’t it kind of ironic that you think you are some kind of expert about the risks of vaccines, yet you don’t understand a concept as simple as false equivocation? Oh wait, that’s not ironic at all….
@Douglas Watts
You are full of crap. I have never read here (in Maine, which I unfortunately just learned you live in, too) about farm animal abuse. I live near a farm and the owner said there are no such laws as you state them. The farmers in Maine, as well as everywhere else, take care of their animals well because they want healthy animals, otherwise they would be useless. And it’s the right thing to do.
One of my former patients is a Maine state veterinarian. She also says you are full of crap.
I have read many stories over the years about animal mills (“puppy mills” or whatever you call them) in Maine. And there have been some ARA that were found to have animals in filthy conditions in shared quarters, animals they “rescued”. They have been justifiably shut down.
By the way, SciBlogs is going berzerk again. For some reason my first comment in this thread showed up under Douglas Watts’ name (Posted by: Douglas Watts | March 1, 2010 4:36 PM). It’s not him who said the following blockquote, it’s me:
Just how good are computers these days? How capable are they? Can modern computers produce an accurate model of animals? Of an organ? Of an animal cell? Of a mitrochondrium from an animal’s cell.
What about a part of the ATP cycle?
Didn’t think so.
The best we can do is model a vague approximation of the subject, because of the inherent limitations of our knowledge and our tools. We don’t know enough, and are tools are too weak, for use to model even as simple an animal as a placozoan. We have a good idea of what a placozoan’s digesting cells do, but we have no idea what their sex cells do. More complex animals are even more mysterious.
We have a tendency to overestimate what we’re capable of, and what our technology can do. Modeling animals and their response to disease is still far off in time, and probably won’t be possible in any meaningful way for generations to come. We don’t even know how much we need to learn to model animals in any meaningful way, that’s how profound our ignorance is.
Just consider the problems our climate models have had. Indeed, we would be far better off if we let the data we’ve gathered speak for itself, rather than trying to speak for it through incomplete and so inaccurate climatic models. Seeing as how our ignorance of climatic systems is so deep, what makes us think we understand any better the vastly more complex systems that make up the simplest animal.
Model cats, frogs, worms, mixozoans? We can’t even model a brief squall beyond the most basic degree, what makes us think our computers, our glorified counting machines can simulate life?
“A STATEMENT OF FACT CANNOT BE INSOLENT.” True. But based on the evidence which purports to substantiate your argument, it most certainly can be WRONG, making it in fact, not factual.
And to the guy who said that animals weren’t sentient creatures–have you looked up the word? Um… ’cause in case you didn’t know, they absolutely ARE sentient. Dumb fuck.
And who or what is this Orac (aside from the description on the website)? So you’re a surgeon who believes in animal testing? Big fucking deal. There are lots of surgeons who don’t. In fact, there are many physicians who don’t believe that animal testing is necessary– http://www.pcrm.org (Physicians Committee for Responsible Medicine)–should anyone on the site care to investigate.
The fact of the matter is, that while animals are LIKE us, they are NOT us–they are NOT human. And as close as you can come in the laboratory to simulating human efficacy of drugs, research or whatever–it’s still a SIMULATION. It’s your best guess.
Rather than use animals, use humans. Prisoners come to mind. Pedophiles. Murderers. Rapists. Animal Abusers. Any one of them would make a fine test subject, and a more accurate and reliable one simply based on biology and physiology. Now this, unequivocally, is FACT.
Sad that a person who operates dog fighting rings has more rights than the dogs he abuses. Even sadder that in the name of “science” (and I use that term extremely loosely) the atrocities committed against animals are even worse and rights aren’t even an issue. Did anyone ask the animal whether or not it wanted to be part of an experiment? Or is it because they can’t speak English or stand upright or have opposable thumbs that you think you are superior to them? I think the last statement is true. And you are so wrong in your thinking. It’s very sad, actually. You are only superior to them in that IT IS YOUR RESPONSIBILITY to protect them. While in captivity, or domesticated, they are completely dependent upon humans–upon YOU– to take care of them. When abuse occurs instead of care, when hurt is inflicted instead of love given and when trust is betrayed, that’s not being human or humane. It’s just plain wrong.
I’m wish I’d never come across your site. It is so sad to me that people like you exist.
I’m wish I’d never come across your site.
…That makes two of us. Thanks for the pile of self-righteous, ethically-challenged excrement. Don’t let the internet door hit you on the way out.
Jillian
Please provide evidence that mice are sentient.
Please provide an ethically defensible argument that whilst it is not OK to experiment on a mouse it is OK to do the same experiment on a human (prisoner, dog fight operator, rapist or whatever other monkier you use to devalue a human life).
In constancy with your stance please stop using anything derived from biological science. This includes shoes and healthcare.
Thanks to all of you for your brilliant posts–it was a delight to follow this thread! I’m going to send the original blog and some of the more salient comments to all my friends who will undoubtedly become regular lurkers on the off chance some other PeTAphile will arrive and throw down the gauntlet to another round of ‘moral superiority’.
Who says science ain’t fun? Hee-hee!
“Rather than use animals, use humans. Prisoners come to mind. Pedophiles. Murderers. Rapists. Animal Abusers. Any one of them would make a fine test subject, and a more accurate and reliable one simply based on biology and physiology. Now this, unequivocally, is FACT.”
Since it’s a well known FACT that the justice system is 100% accurate and never has any biases or the like this is a great plan.
Just kidding. Let us contemplate that the innocence project (http://www.innocenceproject.org/) has exonerated 251 people based on DNA evidence.
You’re suggesting that medicine be tested on potentially innocent people? I wish you’d never come across this site. It is so sad to me that people like you exist.
Yes, I know I’m late to the party, but mice absolutely are sentient. What they aren’t is sapient. Many people use treat the two words as synonyms although they are not. Sapience requires self awareness, while sentience only requires some degree of environmental awareness beyond simple stimulus-response, i.e. the ability to feel.