DNA电脑对付HBV也许有用 DNA computer

as100200

对DNA有反应的生物分子电脑未来也许可用于对付HBV。

DNA computer could fight cancer
By Charles Choi
United Press International
Published 4/29/2004 9:00 AM

NEW YORK, April 29 (UPI) -- New computers made of biological molecules that react to DNA hold the promise to diagnose and treat diseases such as cancer by operating like doctors inside the body, Israeli scientists said.

The devices, used in test-tube experiments, already have demonstrated the ability to identify and then destroy prostate and lung cancer cells, but their creators cautioned it could be decades before such biological computers find their way into medicine.

"The hope is that someday this direction will help lead to a new concept of 'smart drugs,'" said researcher Ehud Shapiro, a computer scientist at the Weizmann Institute of Science in Rehovat, Israel.

"Today's drugs are like carpet bombing -- you take a large amount of drug molecules and they go all over the body," Shapiro told United Press International from Brussels. "A smart drug would only operate where the condition of disease holds."

Shapiro and colleagues have created a series of biological computers in the past five years -- several trillion of which can fit in a drop of water. Their software is made of DNA, while the hardware is made of DNA-manipulating enzymes.

The computers work on genetic material, specifically ribonucleic acid or RNA, the smaller cousin of DNA. The body uses RNA often to transmit messages in the cell. Strands of DNA and RNA can bind if the sequences of molecules that make both up match.

Diseases such as cancer leave their own chemical fingerprint in the body, including over-producing or under-producing specific RNA sequences. The computer's enzyme hardware chops up the RNA it finds. If those bits bind to a computer's DNA software -- which is encoded to look for cancer sequences -- the computer then can release a drug.

In the lab experiments at the Weizmann Institute, the drugs in question were short DNA molecules known to interfere with tumors by suppressing key cancer genes, making those diseased cells self-destruct.

"I think it's really a wonderful landmark piece of work. It certainly merits a lot of interest," computer scientist John Reif at Duke University in Durham, N.C., told UPI. "The goal would be to do this in a cell, to have a doctor in a cell. This experiment doesn't do that. It does it in a test tube -- sort of a doctor in a test tube -- so it's a really cool idea but there's a major hurdle next.

"This is the first step, the biggest and most significant piece of work to date that indicates DNA computing techniques can be really important to therapeutic medicine," Reif said.

The scientists presented their findings Wednesday in Brussels at a symposium where Nobel laureates discussed the future of the life sciences. Their data also appears in the April 29 issue of the British journal Nature.

"We're sort of running ahead of ourselves. Had you asked me a year ago when we started how long it would take to reach the milestone we reached today, I'd have said 10 to 15 years," Shapiro said. "We are still overwhelmed by what we achieved. It took us less time than we thought."

The research team ensured more than just one chemical marker of a disease is required to activate drug output from one of the microscopic computers. A lone symptom could just be a temporary, insignificant phenomenon in the body.

"In our particular design, we can diagnose the presence of a disease by looking in a relatively straightforward way for eight symptoms of a disease, and can go up to 16," Shapiro said.

Shapiro's team originally designed biological computers to compete against electronic computers. The field began in 1994, when computer scientist Len Adleman at the University of Southern California proposed how DNA could be used in solving certain important mathematical calculations, such as the so-called "traveling salesman problem," critical in planning any kind of deliveries in a complex network, from shipping freight to scheduling airline flights to transmitting data packets on the Internet.

The problem is that although a single drop of water can have trillions of biological computers working on a single problem, they moved slowly compared with electronics "and unreliably also," Bennett explained.

Because the biological computer concept did not look as if it could vie with electronics on general computing and win, Shapiro said he decided to "go back and do something useful with it."

"Electronic computers can talk very easily with other electronic devices -- a printer, a DVD player -- but they cannot talk very easily with biological molecules. We opened with biological computers instead to talk with biological molecules," he explained.

The new computers still have a long way to go before they can find use in the body, Shapiro cautioned.

"What we demonstrated is just small enough and smart enough to do the job in a test tube. Making sure it works inside a tissue culture, let alone a living organism, is going to be a challenge," Shapiro said.

"They're not packaged right now," he added. "They'd probably degrade too soon in the body to be active or effective. They'd have to be protected in some way. Also, the enzyme we used here as hardware, Fok1, cuts DNA. If it finds its way into the nucleus of a cell, it'd just chop the entire genome to pieces, and that particular cell would not be too useful afterwards," Shapiro said.

"It's a very cute idea to have chemicals within the system you're working on doing computations on that system," said physicist Charles Bennett of IBM Research in Yorktown Heights, N.Y. "Instead of having to stick a probe into a mixture of cells to find out what's happening inside, these have the advantage of having computations right in there. The thing that's doing the measuring is already part of the mixture."

Bennett also suggested expanding the range of molecules the computer could process and the types of drugs it could make.

The Weizmann team envisions future versions that could even release proteins or other compounds.

--

Charles Choi covers research for UPI Science News. E-mail [email protected]



Copyright © 2001-2004 United Press International

天狼星

DNA计算机开始显示雏形 能够在人体内诊治疾病
--------------------------------------------------------------------------------
http://tech.tom.com　 2004年04月29日 08:47??来源:硅谷动力??月牙儿编译


【eNews消息】科学家已经向制造极其微小的DNA计算机靠近了一步，终有一天，它能够在人体内发现象癌症这样的疾病，并释放出治疗疾病的药物。


　　埃胡德教授以及以色列魏兹曼学院的研究人员在数年前就建造了最小的生物分子计算机，现在，在实验室的实验中，他们已经能够使它分析生物信息，发现和治疗前列腺癌和肺癌。埃胡德说，我们已经给它增加了输入/输出系统，它能够诊断出疾病，并在试管中制造出相应的药物。


　　这种计算机的尺寸非常地小，一滴水中就可能包含有1成亿个计算机。它的输入/输出模块以及软件都是由DNA分子构成的。


　　这一技术能够给癌症等疾病在未来的诊治带来革命性的变化，无需再进行切片检查，DNA计算机能够在人体内的组织中诊断疾病。埃胡德说，我们的医疗计算机可能被看作一种药物，由血液带到全身的各处，检查每个细胞是否已经发生了病变。


　　它能够使医生在肿瘤形成前治疗癌症，如果疾病已经扩散到身体的其它部分，它会向“顽固的”细胞释放药物。不同的输入模块能够诊治不同的疾病。


　　埃胡德承认，要制造成功在人体内诊治疾病的DNA计算机还有漫长的路要走。他说，这还存在许多困难，可能还需要数十年的时间。

[此贴子已经被作者于2004-5-1 8:42:43编辑过]

天狼星

DNA电脑与生物电脑之父

——雷纳德·阿德勒曼

1994年，美国南加州大学教授雷纳德·阿德勒曼（L.Adleman）博士，在《科学》杂志上发表一篇题为《组合问题的生物电脑解决方案》的论文，首次提出分子计算机，即用DNA分子构建电脑的设想。作为一位理论数学家，阿德勒曼教授的研究课题十分广泛，他曾与别人合作发明了用于通信的RSA加密码，RSA中的“A”就是他姓名的首字母；此外，他的研究触角也涉及到了爱滋病和生物学领域。

DNA（脱氧核糖核酸）是生物基因的物质实体，携带着生命信息的密码。过去大多数科学家都认为，以DNA分子而非电荷形式来处理数据，不过是一种有趣的理论化概念，在几十年之内将不会得到实际的验证。然而，阿德勒曼偏偏做成了这种不可能的“芯片”。以目前的情况，叫它“芯片”或许有点牵强，因为它不再是固体，而是由装满有机液体的微型试管组成，承担计算任务的DNA分子就溶在液体之中。

更令电脑界惊讶的是，阿德勒曼竟然利用他发明的DNA生物电脑，解决了一个实际的数学难题。这个题目是这样的：“由14条单行道连接着7座城市，请找出走过上述全部城市的最近路途，而且不能走回头路。” 学数学的人都知道，这是一个经典的数学问题，又叫“推销员问题”（它的正式的名称叫“汉密尔顿路径问题”），该问题的叙述是这样的：“如果一个推销员要在许多个城市推销，每个城市必须而且只能经过一次，如何找到最短的路程？”经典数学中并没有公式可以回答，惟一的解决办法是找到所有可能的路程加以比较，选出最短的一种。然而，即使仅有四个城市，推销员也已面临着12种选择，当然比较所有的路线仍有可能，但随着城市数目的增加，路径将呈现指数增长，穷尽所有的路径变得越来越不可能。

阿德勒曼教授设法驱使试管中的DNA分子来完成计算，他用DNA单链代表每座城市及城市之间的道路，并顺序编码。这样一来，每条道路“粘性的两端”就会根据DNA组合的化学规则，与两座正确的城市相连。然后，他在试管中把这些DNA链的几十亿个副本混合起来，让它们以无数种可能的组合连接在一起。其基本工作原理是：单条DNA以预定的方式和与之对应的DNA相配接。通过7天时间的系列生化反应，DNA电脑自动找出了解决问题的唯一答案，即只经过每座城市一次且顺序最短的DNA分子链。这就是说，用生物学方法模拟的逻辑运算，用一个星期时间完成了电脑几年才能完成的工作，表明了用DNA技术处理高难度数学问题的巨大潜力。

阿德勒曼的成功实验表明，DNA生物电脑已经不是什么科学幻想，它不但打破了传统意义上的计算机概念，而且有助于揭示生命的本质。阿德勒曼发表的那篇报告，不仅使他成为“生物电脑之父”，而且促使了世界各地至少12个实验室开始这方面的研究工作。现在，许多科学家都把DNA计算技术看作是可以用来替代传统电子技术的主要候选者。

例如，美国新泽西州贝尔实验室的研究者，正在改进阿德勒曼的第一台DNA电脑。该实验室的物理学家艾伦·米尔斯说：“我们所做的不再是连接线路，而是将大量的DNA装入罐子，加进盐和酶，然后加以培养，于是不同的 DNA分子就开始互相寻找配对。” 米尔斯的目标是制造一台神经网络计算机，基本构造与人的大脑相仿，不像传统计算机那样采用数字输入，而是一种类比输入。麦迪逊威斯康星大学的研究小组却采取了不同于阿德勒曼的试管办法，他们把DNA链固定到一块镀金的玻璃载片上，使之真正成为DNA芯片。在经过数年的研究之后，该研究小组制造出了几台DNA芯片计算机，每台都由大约100万亿个由人工合成的DNA链状结构组成。英国利物浦大学的马丁·科莫斯等科学家则希望把DNA计算技术送回到活细胞中，在转基因细胞内部模拟计算机逻辑电路，打算在细菌细胞中实现生物“开关”元件。

DNA生物电脑的最大优点，还在于它惊人的存贮容量和运算速度。纳米技术家认为，DNA具有在极小空间里存储海量信息的自然特性，遗传密码符号的间距仅有0.34纳米，1立方米的DNA溶液可存储1万亿亿比特数据；1立方厘米DNA溶液将超过1万亿片CD光盘的存储容量。具有生命特征的这种电脑，运算次数甚至可以达到每秒10的20次方或更高，消耗的能量却微不足道，只有普通电脑的十亿分之一。据说，十几个小时的DNA计算，就相当于人类社会所有电脑问世以来的运算总量。我国国家智能计算机研究开发中心主任、主持研制“曙光”超级电脑的李国杰院士提出，生物计算机要成为一种通用计算机，必须先建立与图林机类似的计算模型。现在DNA电脑最大的问题是很难检测计算结果，一旦这个问题得到解决，DNA生物电脑（芯片）将很快进入实用阶段。

据报道，2001年11月，以色列科学家已经成功研制出世界上第一台可编程DNA电脑，这种电脑即使有一万亿“台”，其体积也不超过一滴水的大小。然而，如何真正替代硅芯片成为普遍使用的DNA微处理器，科学界仍然面临着许多挑战。DNA链的并行处理能力非常适合解决类似“推销员问题”，但随着问题复杂程度的增加，DNA数量也将呈几何级数上升。如果推销员要走遍200个城市，生物电脑所需要DNA分子的总量甚至会超过地球的重量。因而，有些专家更倾向于一种“杂交”电脑，让硅芯片和DNA芯片共同承担计算任务。

阿德勒曼教授曾经说道：“我并不期待构建一台像PC机那样的DNA电脑，但是，生物电脑可以做用其他技术所不能完成的工作。”他预言说，到2002年DNA电脑就可以解决有20个变量的数学问题。DNA电脑将采用其本身的“语言”，以四进制系统来编码，与“人工生命”的研究范畴将融合在一起。对此，阿德勒曼认为，今后的工程技术人员应该接受更加广泛的科学教育，使自己成为“通才”，全面掌握数学、物理、化学、生物学和计算机科学知识，才能做出更多的发明和创新。或许，这正是他本人的切身体会。                          （叶  平）