Identification of high affinity aptamers using massively parallel DNA sequencing

绝大多数的分子太小了，用肉眼是看不见的。在大多数情况下，检测它们的唯一方法是使用其他分子，就像钥匙适合锁一样。人体会产生这样的分子，叫做抗体。当我们被有害细菌入侵时，抗体会粘附在它们身上，并将它们标记出来进行破坏。许多年前，科学家们意识到他们可以使用抗体来检测几乎任何类型的分子。这为医学、医疗保健和研究领域的一系列新技术打开了大门。DNA是大多数人所熟悉的包含生物信息的分子，但它也可以折叠成类似抗体的三维形状。这些结构的形状取决于DNA中编码的信息序列。20年前，科学家们推测，将DNA的信息特性与其识别其他分子的能力联合收割机结合起来是可能的。在一个类似自然选择的过程中，他们将许多不同的DNA序列与一种蛋白质分子混合在一起。大多数人都没有坚持，但也有不少人坚持了，其中一些人比其他人更紧。然后，科学家们利用DNA中的信息制造了许多序列的拷贝，这些拷贝再次将这些扩增的群体与蛋白质粘在一起并混合。这一次，一些在第一轮选择中幸存下来的DNA被更紧密地粘附在蛋白质上的序列排除在外。这些序列被丢弃，而那些粘在一起的序列被扩增，以产生更丰富的群体。经过多轮选择和扩增后，只剩下少数序列。科学家们称这些幸存的序列为适体，这个拉丁词描述了其他分子融入它们的方式，就像钥匙插入锁一样。适体具有许多优于抗体的优点。它们更小，更强大，一旦编码在适体中的信息已知，就可以廉价地大量生产。有了这些优点，人们可能会认为适体很久以前就可以取代抗体，但在发现20年后，它们仍然是弱关系。问题是适体不像抗体那样紧密地粘附在其他分子上，最近科学家们已经找到了原因。用来识别它们的自然选择过程不仅消除了根本不粘蛋白质的序列，而且还消除了粘蛋白质强度低于最强蛋白质的序列。如果这些较弱的序列连接到最强的序列上，就会产生一种新的适体，这种适体与蛋白质的粘附力比原来的强数百倍。读取DNA中编码的信息的技术被称为测序技术。当适体在20年前首次被发现时，需要花费大量的精力来读取单个适体的序列，即使它包含的信息不到100位。相比之下，人类基因组中的30亿比特信息可以在几天内读取。这些进步使得读取所有与蛋白质结合的DNA序列中编码的信息成为可能，而不仅仅是与蛋白质结合最强的少数DNA序列。这就是我们在这个项目中要做的。当我们读取了所有的序列后，我们将使用科学家用来理解人类基因组的相同计算技术将它们组装成一个巨大的表格。这张表将告诉我们序列是如何连接在一起的，以制造一种适体，它能像抗体一样紧紧地粘在分子上。通过制造像抗体一样紧密粘附的适体，我们将打破阻碍其其他优势被利用的障碍。这将导致新的和改进的测试，使科学家和医生能够检测到许多不同种类的分子在同一分钟点的血液，和新的药物，寻找和摧毁癌细胞和有害病毒。

The vast majority of molecules are far too small to be seen with the unaided eye. In most cases the only way to detect them is with other molecules that fit round them like a key fits round a lock. The human body produces molecules like this called antibodies. When we are invaded by harmful bacteria antibodies stick to them and mark them out for destruction. Many years ago scientists realized that they could use antibodies to detect almost any kind of molecule. This opened the door to a whole range of new technologies in medicine, healthcare and research. DNA is familiar to most people as the molecule that contains the information for living things, but it can also fold into three-dimensional shapes that resemble antibodies. The shape of these structures depends on the sequence of information encoded in the DNA. Twenty years ago scientists speculated that it might be possible to combine the information properties of DNA with its ability to recognize other molecules. In a process that resembles natural selection they mixed many different DNA sequences with a single type of protein molecule. Most of them did not stick, but quite a few did, some of them more tightly than others. The scientists then used the information in DNA to make many copies of the sequences that stuck and mixed this amplified population with the protein again. This time some of the DNA that survived the first round of selection was excluded by sequences that stuck to the protein more tightly. These sequences were discarded while those that stuck were amplified to produce an even more enriched population. After many rounds of selection and amplification only a few sequences remained. The scientists called these surviving sequences aptamers after a Latin word that describes the way that other molecules fit into them like a key fits into a lock. Aptamers have many advantages over antibodies. They are smaller and more robust, and once the information encoded in an aptamer is known large amounts of it can be made inexpensively. With advantages like these it might be thought that aptamers would have supplanted antibodies long ago, but twenty years after their discovery they are still the poor-relation. The problem is that aptamers do not stick to other molecules as tightly as antibodies and recently scientists have found out why. The natural selection process used to identify them not only eliminates sequences that do not stick to the protein at all but also sequences that stick to it less strongly than the strongest. If these weaker sequences are joined to the strongest sequence a new aptamer is produced that sticks to the protein hundreds of times more tightly than the original. Technologies that read the information encoded in DNA are known as sequencing technologies. When aptamers were first discovered twenty years ago it required a great deal of effort to read the sequence of a single aptamer even though it contained less than a hundred bits of information. Now by contrast the entire 3 billion bits of information in the human genome can be read in only a few days. These advances have made it feasible to read the information encoded in all the DNA sequences that bind to a protein and not just the few that bind to it most strongly. This is what we will do in this project. When we have read all the sequences we will assemble them into a vast table using the same computing techniques that scientists use to understand the human genome. This table will tell us how sequences can be linked together to make an aptamer that sticks to molecules as tightly as an antibody. By making aptamers that stick as tightly as antibodies we will break down the barrier that is preventing their other advantages from being used. The will lead to new and improved tests that allow scientists and physicians to detect many different kinds of molecule in the same minute spot of blood, and new drugs that seek out and destroy cancer cells and harmful viruses.