Large genome models used to design new viruses

大型基因组模型被用于设计新型病毒

A lot of the AI work in biology has been focused on designing proteins. That’s partly because proteins do most of the business of life, catalyzing the interesting chemistry and structuring cells. So, figuring out how to make a new protein can mean directly tinkering with biochemistry, providing new and potentially useful functions. 生物学领域中，大部分人工智能研究都集中在蛋白质设计上。这部分是因为蛋白质承担了生命活动的大部分功能，它们催化有趣的化学反应并构建细胞结构。因此，弄清楚如何制造一种新蛋白质，意味着可以直接对生物化学进行改造，从而提供新的、潜在有用的功能。

Since the genetic code provides a layer of abstraction between DNA and proteins, it wasn’t obvious what a model trained on DNA could do. Yet people went ahead and made a large genome model, and it turned out to be able to output DNA sequences that could encode functional proteins in bacteria and mimic the gene structures found in complex cells. 由于遗传密码在 DNA 和蛋白质之间提供了一个抽象层，因此基于 DNA 训练的模型能发挥什么作用并不显而易见。然而，人们还是开发出了大型基因组模型，结果证明它能够输出在细菌中编码功能性蛋白质的 DNA 序列，并模拟复杂细胞中发现的基因结构。

Now, those same models have been used to output the genomes of viruses that infect bacteria. This isn’t science fiction—all the viruses the models created are closely related to an existing virus. But they do have some distinct features that would be challenging to evolve. And the researchers who did the work, based at Stanford University, suggest we may want to start thinking now about preparing for the potential that someone could develop a related AI that can design a virus that targets vertebrates. 现在，这些模型已被用于输出感染细菌的病毒基因组。这并非科幻小说——模型创建的所有病毒都与现有病毒密切相关。但它们确实具备一些难以通过自然进化产生的独特特征。这项研究的斯坦福大学研究人员建议，我们现在可能需要开始考虑，为未来有人开发出能设计针对脊椎动物病毒的相关人工智能做好准备。

Large genome models

大型基因组模型

Large language models are essentially trained on their ability to predict the next bit of verbiage in as large a compendium of human-generated text as their developers can get ahold of. Large genome models are the same approach, but applied to DNA. To an extent, that simplifies matters, given that DNA only uses four “letters,” A, T, C, and G. But it’s more complicated in that genomes typically have areas where the next letter matters a great deal, interspersed with sequences where the next base could be anything and it wouldn’t matter. 大型语言模型本质上是通过预测下一个词汇的能力进行训练的，其训练数据是开发者所能获取的尽可能庞大的人类文本合集。大型基因组模型采用了相同的方法，只是将其应用于 DNA。在某种程度上，这简化了问题，因为 DNA 只使用 A、T、C 和 G 这四个“字母”。但它也更复杂，因为基因组中通常存在一些区域，下一个字母至关重要，而这些区域又穿插着一些序列，在这些序列中，下一个碱基是什么并不重要。

So large genome models need to recognize the biological context of the sequence they’re outputting in many situations where we humans haven’t figured it out yet. Yet, if we feed them enough genome sequences, they seem to be able to. Bacterial genes with related functions tend to cluster together. Prompt a large genome model with the sequence of part of a cluster, and it’ll output DNA sequences that encode proteins with related functions—potentially including working proteins that look like nothing else we’ve identified so far. 因此，大型基因组模型需要在许多人类尚未完全理解的情况下，识别其输出序列的生物学背景。然而，如果我们向它们输入足够多的基因组序列，它们似乎能够做到这一点。具有相关功能的细菌基因往往聚集在一起。用一个基因簇的部分序列作为提示词输入大型基因组模型，它就会输出编码具有相关功能蛋白质的 DNA 序列——这可能包括我们迄今为止尚未发现的、功能完备的蛋白质。

But our own lack of knowledge places limits on what we can do with these models. If we prompt them with a bit of sequence from a complex cell, it will respond with a string of bases that contain what look like genes and regulatory DNA. But, since most genes can be located essentially anywhere in a eukaryotic genome, we have no idea what functions these hallucinated genes might perform (if any), so we can’t really test them in any way. 但我们自身知识的匮乏限制了我们利用这些模型的能力。如果我们用来自复杂细胞的一小段序列作为提示，它会回应一串包含看起来像基因和调控 DNA 的碱基。但是，由于大多数基因基本上可以位于真核基因组的任何位置，我们根本不知道这些“幻觉”出来的基因可能执行什么功能（如果有的话），因此我们无法以任何方式对它们进行测试。

Still, as a precaution while training these models, termed Evo 1 and Evo 2, the researchers did not provide them with any sequences from viruses that target complex cells. Even if we can’t understand what they output, there’s a chance that they’ll output something dangerous. But there’s a whole world of viruses that attack bacteria and can’t infect humans. And those, the Evo developers figured, are fair game. So, rather than looking at whether Evo 1 and Evo 2 could output reasonable-looking genes, they decided to test if they could output an entire genome. 尽管如此，作为预防措施，研究人员在训练这些被称为 Evo 1 和 Evo 2 的模型时，没有向它们提供任何针对复杂细胞的病毒序列。即使我们无法理解它们的输出，它们仍有可能输出危险的东西。但世界上存在大量攻击细菌且无法感染人类的病毒。Evo 的开发者认为，这些病毒是可以研究的对象。因此，他们没有去观察 Evo 1 和 Evo 2 是否能输出看起来合理的基因，而是决定测试它们是否能输出完整的基因组。

Order from chaos

混沌中的秩序

The model virus they chose as a test case is the catchily named ΦX174, part of a larger family of viruses that infect E. coli. In addition to the convenience of using the best-studied bacteria on the planet for tests, ΦX174 has a number of notable features. It’s fairly simple: At 11 genes spread over about 5,400 bases, all 11 of those genes have identified functions, and the virus’s infection cycle is well characterized. 他们选择作为测试案例的模型病毒是名字响亮的 ΦX174，它是感染大肠杆菌的庞大病毒家族的一员。除了使用地球上研究最透彻的细菌进行测试非常方便外，ΦX174 还具有许多显著特征。它相当简单：仅有 11 个基因分布在大约 5,400 个碱基上，所有 11 个基因的功能都已明确，且该病毒的感染周期也得到了充分表征。

It also has a useful feature from the perspective of working with a large genome model: The end of the virus always has the same short sequence of bases. So, if you use those bases as a prompt, the large genome model should be able to recognize that it needs to respond by outputting a sequence that is in some way related to ΦX174. To further prepare their models, Evo 1 and Evo 2 were fed over 2 million additional bases of DNA sequences from viruses that infect bacteria (termed bacteriophages), and then fine-tuned with sequences specific to Microviridae, the group that ΦX174 belongs to. 从大型基因组模型的工作角度来看，它还有一个有用的特征：病毒末端总是具有相同的短碱基序列。因此，如果你使用这些碱基作为提示，大型基因组模型应该能够识别出它需要通过输出某种与 ΦX174 相关的序列来做出响应。为了进一步训练模型，研究人员向 Evo 1 和 Evo 2 输入了超过 200 万个额外的感染细菌的病毒（称为噬菌体）的 DNA 序列，然后使用 ΦX174 所属的微小噬菌体科（Microviridae）的特定序列对模型进行了微调。

After that, they experimented with different prompts. Prompt with too much of the ΦX174 start sequence, and it would simply spit the rest of the genome back out. Too little, and it would generate lots of unrelated sequences. The researchers eventually found prompts with four to nine bases of the start sequence worked best. While the outputs were primed to return something that looked like ΦX174, they literally could be anything—from near-carbon-copies of the normal virus to sequences that are only vaguely viral. 此后，他们尝试了不同的提示词。如果提示词中包含过多的 ΦX174 起始序列，模型只会简单地吐出基因组的其余部分。如果提示词太少，它会生成大量不相关的序列。研究人员最终发现，使用起始序列中 4 到 9 个碱基作为提示词效果最好。虽然输出结果被设定为返回看起来像 ΦX174 的内容，但它们实际上可能是任何东西——从正常病毒的近乎复制品到仅有模糊病毒特征的序列。

To address this, the research team put a lot of pre-conditions on the outputs designed to throw away some of the more extreme ones while still enforcing a bit of tinkering. For example, ΦX174 uses its own version of a spike protein to latch onto and infect bacteria; if the gene that encodes spike is missing or damaged, the virus simply wouldn’t work. So they discarded any outputs that had a spike protein that was less than 60 percent identical to the real one. They also threw out viruses that were too long or too short (< 4,000 bases or > 6,000), any that had strings of the same base more than 10 bases long, and any that had unusual frequencies of the two base pairings (GC and AT). 为了解决这个问题，研究团队对输出结果设置了许多预设条件，旨在剔除一些极端情况，同时仍保留一定的变异。例如，ΦX174 使用其特有的刺突蛋白来附着并感染细菌；如果编码刺突蛋白的基因缺失或受损，病毒就无法发挥作用。因此，他们丢弃了任何刺突蛋白与真实病毒相似度低于 60% 的输出结果。他们还剔除了过长或过短（小于 4,000 或大于 6,000 个碱基）的病毒，以及任何包含超过 10 个相同碱基连续排列的序列，以及任何两种碱基配对（GC 和 AT）频率异常的序列。

Throwing out those at the computer analysis stage left them with a reasonable number of potential outputs to test: 302 proposed viral sequences. They were able to chemically synthesize the sequences of 285 of them and inserted them into bacteria to see what happened. In most cases, the answer was “nothing.” But 16 of the sequen 在计算机分析阶段剔除这些序列后，他们剩下了一定数量的潜在输出结果可供测试：302 个拟议的病毒序列。他们成功化学合成了其中的 285 个序列，并将它们植入细菌中以观察结果。在大多数情况下，答案是“什么也没发生”。但其中 16 个序列……