现代遗传学

Logic Model (结题作业) Template:LogicModel 11 1, 2021
Abstract Specification Context 接下来将以遗传学为例探究学科认知的共同性。遗传学是研究生物代际信息传递的学科。遗传学在历史上共发生了三次重大的范式转变。第一次是从前范式时期到孟德尔遗传学；第二次是从孟德尔的颗粒遗传学到经典遗传学；第三次是从经典遗传学发展到与分子生物学、生物化学结合的现代遗传学。 Goal 蒋玉骅“超越学科的认知基础”课程主页 课程感悟 科学革命的结构读后感 Success Criteria 第一：范式所揭示的事物之本质 第二：虽普遍但较少的事实判定（边界条件） 第三：与实证数据相匹配的范式 第四：选择问题的标准（可建模） 第五：规则的抽象化（通用性） 第六：解答谜题所需的空间与时间
Concrete Implementation Given Inputs When Process is executed... Then, we get Outputs 1、《科学革命的结构》这本书会使我对学科的发展有新的认知； 2、在课程中所学习的新的概念和新的思维方式也将会使我对跨越学科的问题有新的认知。 遗传学的前范式时代 孟德尔遗传学 经典遗传学 现代遗传学 1、设计个人主页； 2、Final Project 3、《科学革命的结构》读后感
Boundary/Safety Conditions of 结题作业 1、有时课堂上所学的一些概念难以在短时间内准确理解； 2、有时难以看懂老师提供的英文资料，很难明白想要表达的意思。

Final Project

开题报告

理想的偏序集

Error creating thumbnail: Unable to save thumbnail to destination

沃森和克里克的“量子”——遗传因子

   作为一个复杂程度极高的系统，遗传学在孟德尔之前并未形成确定的范式，本质原因是因为没有人找到这个复杂系统的度量衡单位。孟德尔定义了遗传学的“量子”——“遗传因子”，自此遗传学成为了一个可表达（Representable）的系统。在孟德尔的会计方式下，每个可证明的结论都是正确的，从而具备了可靠性（Soundness）；每个正确的结论都是可被证明的，从而具备了完备性（Completeness）

孟德尔的“量子”——遗传因子

   作为孟德尔的话语体系中的“量子”，“遗传因子”扮演了计算机的话语体系中的比特的角色，“遗传因子”有显性和隐形，对应了比特为0或1。但问题随之出现：这样的语言所建立的偏序集并不能趋近遗传系统的各类结构。

   于是才产生了科学革命，“量子”的转变标志着范式的更替。自1953年沃森（J.D.Watson）和克里克（F.H.C.Crick）发现了DNA的双螺旋结构后，遗传学的“量子”由遗传因子变成了碱基对。令人惊奇的是，新的“量子”碱基对也恰有两类：“A-T”“C-G”碱基对的相互影响作用构成了基因这个偏序集，而遗传学的重心变成了破译函子——将基因映射到表现型的函子。这被后人称为中心法则。

简介

   遗传学系统模型的优化是一个反复迭代的模型空间搜寻过程，这在前范式阶段得到了充分的体现。在孟德尔之前，遗传学并未形成确定的范式，而是处于前范式时期。对遗传的认识主要来源于对自然界的观察、农业实践和畜牧业实践。这一时期，相竞争的范式主要是模子学说、蓝图学说、融合遗传学说（Blending inheritance）与获得性遗传学说。这四种学说的产生时期不同、关注点也不同，体现了前范式时期较为典型的特征。
   孟德尔(Mendel)的出现增加了遗传学知识的可表达性，甚至是遗传学计算的连续性（即使使用了类似箭头的符号）。孟德尔开创的的遗传学知识都可以使用相应的箭头来进行描述,从而构建相应的框架.使逻辑模型可进行知识表达(详见[1])。

遗传学的前范式时代

    从范畴学的角度看，遗传学系统模型的优化是一个反复迭代的模型空间搜寻过程，这在前范式阶段得到了充分的体现。模子学说是由希波克拉底提出的，他认为后代的成长就是在逐渐接近父本的模子。然而，亚里士多德基于他的理念哲学和生活观察，提出了蓝图学说，即父代与子代有发展成相同形态的能力。获得性遗传由拉马克提出，是为了解释地层中发现的物种演化现象。融合遗传学说则是由达尔文提出。1859年达尔文在著作《物种起源》一书中提出了以自然选择为中心的进化学说，确立了变异、遗传和选择是生物进化的基本因素。其中的遗传指的就是融合遗传。然而达尔文并不了解三者之间相互联系的机制，更无法回答变异的来源和维持机理。《物种起源》发表后，生物进化理论虽然很快被学术界所承认，但不久又遭到种种非难，其中主要是所谓杂交淹没效应。
    1867年英国数学家兼工程师詹金，从融合遗传理论出发，根据简单的数学计算，认为如果以突变的形式产生新的变异，因其数量有限，单个突变个体同大量正常者交配，以后便会淹没在正常个体里，变异逐渐减弱，终至完全消失，使自然选择无能为力。因此通过选择累积作用而产生进化是难以置信的。詹金对《物种起源》的这种批评使达尔文十分烦恼，穷于应付。
    事实上达尔文创建了一个可表函子,也就是一种可以穿透多层次系统的数据映射机制。在那个时代，达尔文的融合遗传理论十分流行。人们把遗传物质猜测为液体状态的东西，认为任何一个个体都是双亲液体遗传物质相互融合的产物。两种液体一旦融合，各自所具有的特性均被冲淡，最后使变异丧失净尽。对此达尔文也将信将疑。为了让自己的进化学说与融合遗传理论相互协调，他不得不改变原来对突变的看法，主张必须拥有大量的一定变异（不遗传的变异）即大量个体之间的差异，方能给自然选择提供材料，物种才得以进化。同时达尔文在《豢养和栽培下动植物的变异》一书中增加了《外界条件的直接作用和一定作用》、《器官的训练与不训练》两章，同时提出泛生说以解释获得性状的遗传现象。达尔文不懂遗传学的内在规律，不明白遗传在物种进化中的作用而改变了自己的观点，其后果是严重的。此后，遗传学家契特维雷科夫指出：这一改变使他对基因突变作用和意义的看法与现代遗传学距离越来越远；这一改变使他更加接近现代拉马克主义。根源在于没有吸收同时代人的科学成果，反而坚持融合遗传理论。杜勃赞斯基认为要是达尔文“放弃了融合遗传理论，他就把阻碍进化思想前进的道路打通了”。吸收和放弃的可能性与条件都是存在的，然而对达尔文来说，都没有成为现实。

沃森和克里克的贡献

     双螺旋结构是一个包含迭代（iteration）和分支（Branch）的简单结构!(迭代与时间有关，分支与空间有关)，这大大增强了遗传学的可表达性。事实上，结构的复杂性确定更新的时钟速度。

课程感悟

第二周：

 key-value pair 揭示了复杂网络系统工作原理的本质。用这样的方式去理解范式转换，就会发现科学革命实际就是key-value pair 的扩增与删减；不可通约性来自命名空间的不同。而这样的不可通约正是历代科学家致力于消除的，这就像科学家们企图寻找科学语言的“罗塞塔石碑”一样。以人工智能为例：微软近期就提出了深度学习框架的通用语言——repo1.0，他们希望通过构建这一深度学习框架“罗塞塔石碑”，让研究者们能够在不同框架之间轻松运用专业知识，并且实现不同开源社区之间的合作。如果这能实现，就可以创建一个跨语言对比的常用设置（Python、Julia、R），实现不同开源社区之间的合作。
 另外，我个人认为key-value pair也并非全无缺陷。尽管key-value式的查询方法有查询速度快、存放数据量大、支持高并发等等优点，但缺点也是显著的，那就是不能进行复杂的条件查询。甚至当这样的pair数量累计到了一定层次后，key本身的搜寻都极为不易，原本的key也变成了value，需要找到更高一阶的key与之对应才能解决这一问题。

第三周：

 课上的视频让我想到Paul Nurse的一句话：生命的深层结构指的是细胞或有机体这样的基本单元，它们能够自我繁殖，并允许微小的变异。繁殖与变异共同通过自然选择推动物种的演化，从而形成多样化的生物种群。它们不仅能够在变化的环境中存活，还能够利用新的机会。而那些成功适应环境的单元就能继续繁殖后代。
 类似的机制在不同的尺度上都发挥着作用，构成了众多关键生物过程的基础。一个复杂系统会有一些很小的东西组成，由这些很小的东西进行互动。胚胎从单细胞发育为成熟有机体的过程中，会经历好几个生长阶段（人类有几十个），每个都与前一个略有不同。最终，受精卵繁衍出各种不同的细胞，包括心脏、肝脏和脑部的细胞。
 这样的原理成为了冯·诺依曼一些设计的出发点。他曾精确地设计了一个被称为“通用复制器”的数学模型。它包括三个基本组成部分 ：机器A是一台可以根据指令整合资源并进行组装的机器 ；程序B能够指挥机器A ；主程序C可以指挥A来制造A+B+C。从技术上讲，它是一台元胞自动机，可以从周围随机散落的碎片中获取零件。原则上，根据他的设计，我们可以用现代技术造出一个3D打印与计算机的混合系统，它能够收集材料来制作你想要的东西或复制其自身。通过精心设计故意犯错的程序或宽松的质量控制，我们也能解锁生命的另一个秘密——变异。

第四周：

 Composition作为单子的反差，描述的是故事的结构，是复杂系统的总体。水波先分后合的这种形状的变化能够启发一个5岁的小孩（Ted Nelson），并促成他在数十年后创造改变世界的网络系统。所有语言只有4种领域：algebra, boolen,key-value,composition.这4种领域支撑起了复杂结构的所有信息。 用隐喻来明确如何进行超越学科的认知，这样不同学科之间也可以对比。通过命名可以加强人物地名的功能性设定。函数的对象是单个对象，函子对象是系统，自然转换则是比较比较的方法。在函数映射中一个单一的箭头可以分化成若干个压缩前的千丝万缕的箭头。在函子映射中，一个箭头包含了对元素的映射和对元素关系的映射。修订这样映射规则的过程，就是产生更新的范式。
 课程上我们讨论了这些问题：
 （1）函数：A是阿斗，B是刘备。那么A到B，阿斗和刘备的关系是什么。答:刘备是阿斗的爸爸，刘备把位置让给了阿斗。
 （2）函子：S为孙吴，T为曹营。比较S和T之间的关系：答: 孙吴系统犯下了曹魏系统犯过的错
 （3）自然转换：S为创业，T为统一大业，S到T的不同路径进行一下比较。答:从创业到统一或者把所有竞争者都干掉，或者是使用吞并的方法，把所有的敌人都变成自己的人。相互比较后发现，第二种方法更加的和谐。如：曹操使用挟天子以令诸侯的方法来实现统一大业，而其他的政治集团有的使用强行征战的方法，曹操的方法显得更加的阴险，而其他的政治集团的方法显得更加的光明正大。  
 范畴论：水平垂直两种截然不同的复合方式，horizental composition vertical composition。然后讲到了一个图，左边的图和右边图将保持相似的结构，自我感觉范畴论其实就是函数的相互映射，以及映射的映射。越到上层压缩性会变得越高，于是变得更加的抽象，可以使用越来越复杂的矩阵来将之进行运算。对于一个无限复杂的工程问题可以通过范畴论进行相应的化简。
 范畴论之中最为关键的思路是幺元，把所有的东西都放在矩阵之中进行运算，范畴论会被计算科学界用来作为一个编程的工具，范畴论将会被计算科学界用来简化计算的流程，如何才可以将一个复杂的系统进行简化，当所有的名词在人脑之中放不下的时候，我们可以使用范畴论，将他按照一定的方式进行排列整理，使用维基词典进行整理。

第五周：

 文字、会计法则、对称性、与认知能力的演进彼此相关。文字的发明来自记账的需求，知识的传承建立在可表达性上。文字、账目明细、可比对的结构或数据: 创造可重复表达现象的条件。数据的验证方法: 数据内容的差异性，内容的不可篡改性。最“小”的信息（可表达差异性的符号）单位是量子(Quantum)。命名空间的计量方式是枚举系统结构的方法。复式记账、会计平衡公式、逻辑结构的复合与拆解与近代数学与科学方法相关。这样的表达需要具备可靠性（soundness）和完备性（Completeness）。
 目前，许多学科都可以用相应的箭头来描述，从而构建相应的框架。我们可以利用等式和箭头，建立会计学与物理学之间的联系。在会计学中：
                         A（资产）=L（负债）+SE（所有者权益）
 而在物理学中：
                         H（哈密顿量）=T（动能）+V（势能）
 物理学和会计学在此公式上的惊人相似表明所有学科实际共有一套相同的语言表达体系。

科学革命的结构的读后感

Chapter 2

 Paradigms are theories created by one of the pre-paradigm schools. Though the author gives description rather than explanation of the arduous process of acquiring paradigms, we can conclude that when some theories are universally accepted, and can attract an enduring group of adherents away from competing modes of scientific activity, paradigms that prove able to guide the whole group’s research emerge all of a sudden.
 The emergence of a paradigm contributes to scientific inquiry in seven aspects. First, it makes both fact collection and theory articulation highly directed activities, so scientists no longer explore nature casually or at random. Second, it suggests which experiments will be worth performing so scientists are confident what they are studying is highly relevant. Third, the end of interschool debate ends the constant reiteration of fundamentals. The confidence that they are on the right track encouraged scientists to undertake more precise, esoteric, and consuming sorts of work. Fourth, it transforms a group previously interested merely in the study of nature into a profession or a discipline. Fifth, it marks the beginning of specific classification. Sixth, it creates advanced systems that improve effectiveness and efficiency, in particular, for esoteric work. Above all, it produces scientific community whose members push on to more concrete and recondite problems, and increasingly they report their results in articles addressed to other electricians.
 Ultimately, the emergence of a paradigm is invariably followed by a truth boom, as truth emerges more readily from error than from confusion.

Chapter 3

 Although paradigms have shown to be particularly revealing of the nature, the match between facts and theories is still imperfect, and scientists have to avoid approximations and obtain satisfactory agreements.
 To do this, they need to answer the following questions. First, how to conduct empirical works to articulate the paradigm? Second, what further explorations relative to theoretical works can they make to classify theoretical problems of normal science? Third, how to use existing theories to predict factual information of intrinsic value?
 All their efforts serve for the reformulation of a paradigm: to articulate the paradigm elegantly and logically in mathematics, which is both theoretical and experimental. The desire for acknowledgement and the ambition for fame ensure scientists to pursue the same goal. Driven by anthropic ultimate curiosity about the nature, they are committed to solve the problems after the acquisition of the paradigm. 
 The problems always exist, as a successful paradigm is not, however, to be either completely successful with a single problem or notably successful with any large number.

Chapter 4

  What are characteristics of normal science?
  Normal science shares paralleled characteristics with puzzle-solving. First, it offers challenging problems as puzzles that can test ingenuity or skill in solution and thus drive scientists on. Second, the criterion of these problems is the assured existence of a solution, but on the contrary, has nothing to do with the intrinsic value of their outcome. Third, these puzzles are invariably so attractive that scientists attack them with remarkable passion and devotion, holding the conviction that, if only they are skillful enough, they will succeed in solving a puzzle that no one before has solved or solved so well. It is very much the same thing as a child is immersed in solving crossword puzzles that may not give him any benefit.
  Above all, like puzzle-solving, normal science has rules. In the same way as all the pieces must be used before you solve a jigsaw puzzle, until certain conditions have been satisfied, no problem can be solved. Rules play the role of established viewpoint or preconception that bound the admissible solutions to theoretical problems. Under the influence of rules, scientists adopt the attitude that the results of their researches must fall into a narrow range that the paradigm restricts. As a child must obey rules in his games without fail, all these rules have undoubtedly held for scientists at all times.

Chapter 5

  We can account for the priority of paradigms in four aspects. First, rules are abstracted from their more global paradigms and deployed. However, the existence of a paradigm is independent on the existence of rules and need not even imply that any full set of rules exists.
  Second，paradigms can be determined easily, and they can determine normal science directly without the intervention of discoverable rules or assumptions. They are more binding and more complete when it comes to the definitions of basic concepts. Thanks to paradigms, we can understand what makes a particular problem or solution legitimate almost intuitively.
  Third, in the absence of rules, normal science can still proceed under the guidance of paradigms that direct research. As long as paradigms remain secure, however, they can function without agreement over rationalization or without any attempted rationalization at all.
  Fourth, Explicit rules, when they exist, are usually common to a very broad scientific group, but paradigms need not be. They are quite irrelevant and related to each other to a very limited extent, so they need not stand or fall tegother.
  In a nutshell, rules cannot exhaust all functions of paradigms and it is flexible and subject to change. While paradigms might be intangible and artificial, they have absolute dominance over rules.

Chapter 6

 I want to give my account of crisis at the close of a paradigm.
 The close of a paradigm is noncumulative as the threat it face is abruptly posed by a problem that the paradigm fails to solve. After all efforts to reconcile this problem with the paradigm are in vain, this novelty will be claimed incompatible with the basic assumptions of the paradigm. Crisis then occurs with destructive changes in beliefs about nature. Scientists pronounce their failure in saving the paradigm from breakdown, and feel obliged to desert it. A crisis is an occasion for retooling because the tools supplied by the original paradigm prove uncapable of solving all problems. 
 In my opinion, the use of “crisis” is reasonable in a second way. Scientists run the risk of being wrong when they persuade themselves to make the tough decision to convert their beliefs. Together with science, scientists themselves are passing through a grave crisis.

Chapter 7

 In this chapter, Kuhn discusses response to crisis.
 Kuhn first lists two reasons for doubting that scientists reject paradigms because confronted with anomalies. First is that scientists do not renounce the paradigm that has led them into crisis while they consider alternatives and second is that they can actually devise numerous articulations to eliminate any apparent conflict.
 So why do scientists reject paradigms? He notes that to reject one paradigm goes hand in hand with substituting another. Scientists see puzzles as a source of crisis only when from another viewpoint. To make this possible, an anomaly must usually be more than just an anomaly. Kuhn applies a few examples to illustrate how the transition to crisis and to extraordinary science is evoked. 
 Kuhn claims two effects of crisis are universal and crisis are noncumulative. In the end, a transition from normal to extraordinary research results in scientific revolution, which, from my perspective, is the ultimate response to crisis.

Chapter 8

 Kuhn brings up a question:” Why should a change of paradigm be called a revolution?” and outlines the parallelism between scientific and political revolution.
 First, he describes the feature of a scientific revolution: it is noncumulative and the paradigms involved are incompatible. The striking similarity between the characteristics regarding the process of political revolution and scientific revolution are that they begin with a growing sense that existing systems have ceased to meet certain problems. The awareness breeds dissatisfaction. They aim to change the status quo in ways that are prohibited.
 Second, like competing political institutions, competing paradigms proves to be fundamentally incompatible modes. Paradigmatic differences cannot be reconciled.
 Third, the assimilation of either a new sort of phenomenon or a new scientific theory must demand the rejection of an older paradigm, which would make scientific revolution genuinely cumulative otherwise.
 Fourth, Kuhn refutes logical positivist view, arguing that the logical positivist view makes any theory ever used by a significant group of competent scientists immune to attack.
 In the end, since paradigms cannot settle their differences. It is inevitable that they will talk through each other. The results of debates are determined by external criteria. And this recourse to external criteria most obviously makes paradigm debates revolutionary.