It's because above the gene and molecular level, there's no established high -throughput techniques, and also there's no well defined questions to ask. Another reason is, above gene and molecular level, even at multi-gene and multi-molecular level, it's a complex system. There's no analytical solution for almost all the complex systems, which means prediction is almost impossible. Of course you can always do numerical simulation, but that will be impossible to generate something clean and simple as H-H model or E=mc^2. And H-H model is still the simplest diff equations. As an example, George Church once had a science paper with an MIT computer scientist to develop a high-through put system to understand cell morphology: http://arep.med.harvard.edu/pdf/Bakal07.pdf if you go to its 65 pages Supplement, you will understand how hard it is: http://www.sciencemag.org/content/suppl/2007/06/19/316.5832.175 This is already the collaboration of the most top labs in the world, and this is all they get. Sabastian Seung is gathering 10+ MIT computer scientists to work on the connectome project for Brain, I can imagine a much larger scale than this George Church's work. It's a very long way to go, much longer than our life-span.
复杂系统 is actually a word to use when you don't want to speak a paragraph. It's not hard to understand why something is complex system, if you have several components interacting, the evolution of the whole system is almost non-predictable and too easy to go chaotic. Simply just we don't have analytical solution for almost all the PDEs in nature.
一个数学模型,即使很不完美的,也好过没有数学模型。 这话虽然不是生物学家说的,生物里面也成立吧。你觉的Pauling的“On the Nature of the Chemical Bond”很完美吗?一点也不,它实在是简化的太厉害了。没人会认为 两个小球之间连一个小棍是一个真实原子系统。但是,Pauling的贡献在于一个也许你 认为很复杂的体系,其实是可以用一种简单的方式把整个理论框架建立起来。而他以后 的大部分精细的工作,都只是在修补这个框架里的各种细节。 实验数据,其实并不都是客观的。大部分的实验数据,如果没有一个理论框架的承托, 是比较苍白的。有人是鸟,有人是青蛙。鸟看不到细节,而青蛙过于注重细节。这虽是 分工不同所致,但是鸟和青蛙的工作都是必须的。
Almost everything pins down to the analytical solution of PDEs. for many years, if you got an analytical solution of any equations in Einstein's General Relativity theory, it could lead to at least a Nobel prize, or some discovery like black hole. think of just the analytical solution of Heat Equation leads to the B-S option pricing formula, and became once the standard on wall street. My point is, we can't always rely on the luck to get the analytical solutions, there got to be some new math to resolve all the heck or just tell us we can't, otherwise, I don't see any chance that human will understand the life and brain.
The computer revolution gave us the illusion that we can tackle the complex system by running much faster and doing larger scale simulations. It maybe correct in some sense. But on the other hand, we must remember there's almost no progress on the ability to tackle PDEs and SDEs in nature, although fancy words like chaos are everywhere. It's simply just we don't have the tools.
【在 f*******r 的大作中提到】 : 一个数学模型,即使很不完美的,也好过没有数学模型。 : 这话虽然不是生物学家说的,生物里面也成立吧。你觉的Pauling的“On the Nature : of the Chemical Bond”很完美吗?一点也不,它实在是简化的太厉害了。没人会认为 : 两个小球之间连一个小棍是一个真实原子系统。但是,Pauling的贡献在于一个也许你 : 认为很复杂的体系,其实是可以用一种简单的方式把整个理论框架建立起来。而他以后 : 的大部分精细的工作,都只是在修补这个框架里的各种细节。 : 实验数据,其实并不都是客观的。大部分的实验数据,如果没有一个理论框架的承托, : 是比较苍白的。有人是鸟,有人是青蛙。鸟看不到细节,而青蛙过于注重细节。这虽是 : 分工不同所致,但是鸟和青蛙的工作都是必须的。
O*2
66 楼
赞一个。抽象成系统可以帮助很多design。say,Sarpeshkar。
m*7
67 楼
Agree.
high solution will 2. morphology:
【在 d*****r 的大作中提到】 : It's because above the gene and molecular level, there's no established high : -throughput techniques, and also there's no well defined questions to ask. : Another reason is, above gene and molecular level, even at multi-gene and : multi-molecular level, it's a complex system. There's no analytical solution : for almost all the complex systems, which means prediction is almost : impossible. Of course you can always do numerical simulation, but that will : be impossible to generate something clean and simple as H-H model or E=mc^2. : And H-H model is still the simplest diff equations. : As an example, George Church once had a science paper with an MIT computer : scientist to develop a high-through put system to understand cell morphology:
发育的高度重现性也正说明有牢固的法则在指导这个过程 agree with this... the embryo starts from just a few cells, and become such a complex system. This means at least several things: 1. the initial states is limited (I assume all the initial information is stored above or equal to the molecular level), 2. from a few cells to a whole organism, not sure if the cells contains all information for the organism, seems likely; brain maybe different. 3. the final states is converged. Every human has two eyes and one nose. It' s a very amazing convergence. It's quite amazing how this system could get rid of the noise and handle physical construction.. And seems there must be a new math there, just human haven't invent it, or at least prove we can/cannot invent it.
There is a very good resource online for these topics, William Bialek at Princeton has done a lot on this topic, he has given a short course before. He has pointed to all important papers in this field, enjoy: http://www.princeton.edu/~wbialek/rome/rome_course.htm
【在 d*****r 的大作中提到】 : 发育的高度重现性也正说明有牢固的法则在指导这个过程 : agree with this... : the embryo starts from just a few cells, and become such a complex system. : This means at least several things: : 1. the initial states is limited (I assume all the initial information is : stored above or equal to the molecular level), : 2. from a few cells to a whole organism, not sure if the cells contains all : information for the organism, seems likely; brain maybe different. : 3. the final states is converged. Every human has two eyes and one nose. It' : s a very amazing convergence.
u*d
74 楼
非常感谢!
【在 d*****r 的大作中提到】 : There is a very good resource online for these topics, : William Bialek at Princeton has done a lot on this topic, he has given a : short course before. He has pointed to all important papers in this field, : enjoy: : http://www.princeton.edu/~wbialek/rome/rome_course.htm
我几乎可以肯定你没有看过,或者没有仔细看过HH 1952年那篇经典的文章。而且很不 幸的是,你也根本没有仔细阅读我的原文就急不可耐的跳出来。我自己本人就是专门做 离子通道数学模型的,H-H 那篇1952年的文章是需要经常阅读的,尽管如此,在写前文 之前,为了避免错误,我还特意重新看了一遍。因为你没有仔细看过,所以你的叙述当 中错误太多。我指出其中一点,剩下的等你看完了之后我们再讨论: (1) “电导和跨膜离子浓度差的关系不是假设,是根据扩散定律推倒的。” 1952年的原文当中推导的原句是: "...we shall suppose that the sodium conductance is proportional to the such molecules inside the membrane but independent of the number on the outside. From Boltzmann's principle, the proportion "Pi" of molecules on the inside of the membrane is related to the proportion on the outside, "Po", by Pi/Po = exp(函数)“ 另外,我写这篇原文主要是给非神经生物学领域,甚至是非生物领域内的人看的,所以 写的非常直白。"gating"这种词语除了让其他人看不懂之外,还能有什么作用?
【在 k*****1 的大作中提到】 : 我几乎可以肯定你没有看过,或者没有仔细看过HH 1952年那篇经典的文章。而且很不 : 幸的是,你也根本没有仔细阅读我的原文就急不可耐的跳出来。我自己本人就是专门做 : 离子通道数学模型的,H-H 那篇1952年的文章是需要经常阅读的,尽管如此,在写前文 : 之前,为了避免错误,我还特意重新看了一遍。因为你没有仔细看过,所以你的叙述当 : 中错误太多。我指出其中一点,剩下的等你看完了之后我们再讨论: : (1) : “电导和跨膜离子浓度差的关系不是假设,是根据扩散定律推倒的。” : 1952年的原文当中推导的原句是: "...we shall suppose that the sodium : conductance is proportional to the such molecules inside the membrane but : independent of the number on the outside. From Boltzmann's principle, the
我没有觉得我无中生有编了什么话。你最大的错误,离子通道是电容,我没有误解吧。 关于内外离子,你自己的原话把"on the inside/outside of the memberane",描述成 了“膜内外”离子浓度差,也就是"inside/outside of membrane",这让人直接想到的 是flux equation,所以如果你认为我误解了你的话,是因为你的原话用的词是错的。 文中说的很清楚, Boltzmann's principle描述的是一个假设的,贴在膜内外表面的带 电荷粒子,最后还推断出了其带电荷是6个。它既不是钠离子,也不是所谓膜内膜外离 子,按现在的理解,最接近的应该是离子通道本身的voltage sensor。