DeepMind发布AlphaGenome2,可预测基因突变对细胞的影响
DeepMind’s New AI Is A Gift To Humanity
DeepMind又一重磅基础模型发布,解决基因组学关键难题且直接开源,生物计算领域从业者必读。
Your DNA contains roughly 3 billion letters, but reading them doesn't tell us everything about you. Hidden among them are millions of little molecular switches to control when and where your genes become active. And many diseases begin where the switches go wrong. And DeepMind's new AI technique, Alpha Genome 2, is finally able to predict what these little switches do, but with a twist. Okay, so what is this exactly?
你的DNA大约包含30亿个字母,但读取它们并不能告诉我们关于你的一切。隐藏在这些字母中的是数百万个微小的分子开关,用于控制你的基因何时何地变得活跃。许多疾病正是从这些开关出错的地方开始的。而DeepMind的新AI技术Alpha Genome 2终于能够预测这些微小开关的作用,但有一个转折。那么,这究竟是什么呢?
With alpha genome one, you give it a piece of DNA and you change a letter. Is nothing going to happen? Or could this tiny mutation switch on a gene that drives terrible diseases like leukemia? Or could it alter the genes linked to dangerously high cholesterol? The previous alpha genome 1 AI can predict exactly this kind of effect, which is basically insane. I can't believe that I'm saying this. Meanwhile, my Bluetooth headphones still can't connect reliably.
使用Alpha Genome 1时,你给它一段DNA并改变一个字母。会发生什么吗?还是说这个微小的突变会开启驱动白血病等可怕疾病的基因?或者它会改变与危险高胆固醇相关的基因?之前的Alpha Genome 1 AI可以精确预测这类影响,这基本上令人难以置信。我不敢相信我在说这话。与此同时,我的蓝牙耳机仍然无法可靠连接。
So, there's that. However, there is a problem. The same mutation can have a completely different effect based on where it is, which cell it is in. For instance, your brain and skin contain the same DNA, but they use it completely differently. And alpha genome one struggles to predict these finer differences. Of course, because it's kind of impossible. There are billions of single letter mutations and hundreds of different cell types.
所以,这就是情况。然而,存在一个问题。相同的突变可能根据它所在的位置以及所在的细胞类型产生完全不同的影响。例如,你的大脑和皮肤含有相同的DNA,但它们以完全不同的方式使用它。而Alpha Genome 1难以预测这些更细微的差异。当然,因为这在某种程度上是不可能的。有数十亿的单字母突变和数百种不同的细胞类型。
Unfortunately, that is trillions of possible combinations. Testing all of these in a lab is completely hopeless. So, how on earth would you predict all this from just a sequence of DNA? All this knowledge from a bunch of letters. Well, for example, they tested mutations known to cause dangerously excessive insulin production. And finally, alphaggenome 2 predicted that these mutations switch on the HK1 gene in insulin producing beta cells, but have almost no effect in other pancreatic cells.
不幸的是,这意味着可能有数万亿种组合。在实验室中测试所有这些是完全无望的。那么,你究竟如何仅凭一段DNA序列来预测这一切呢?所有这些知识都来自一堆字母。例如,他们测试了已知会导致危险过量胰岛素产生的突变。最终,Alpha Genome 2预测这些突变会在产生胰岛素的β细胞中开启HK1基因,但在其他胰腺细胞中几乎没有影响。
But wait a second, this problem might be exceptionally well suited for a powerful modern AI. Why? Because we can create input output pairs to learn on. The input could be a piece of DNA from the pancreas and a pancreatic beta cell as a cell type. So what is the output? Well, how active its genes are. But wait, scientists have already measured this in a stupendously large number of cells. So we have tons of training data.
但等一下,这个问题可能特别适合强大的现代AI来解决。为什么?因为我们可以创建输入输出对来进行学习。输入可以是来自胰腺的一段DNA以及作为细胞类型的胰腺β细胞。那么输出是什么?嗯,就是其基因的活跃程度。但等等,科学家已经在一个极其庞大的细胞数量中测量过这一点。所以我们拥有大量的训练数据。
And it turns out we can use that data for things other than serving ads to sell more Nissan. Great. Now if you do that there are two big questions. One is AI powerful enough to learn these interactions and if yes question number two can it even work with cell types it hasn't seen yet? So can it generalize? Can it go beyond the training data? Is that even possible? Dear fellow scholars, this is two minute papers with Dr.
事实证明,我们可以利用这些数据来做除了投放广告以销售更多日产汽车之外的事情。太棒了。那么如果你这样做,会有两个大问题。第一是 AI 是否足够强大,能够学习这些交互关系?如果是的话,第二个问题是:它能否处理它尚未见过的细胞类型?也就是说,它能否泛化?它能否超越训练数据?这甚至可能吗?亲爱的学者们,这是 Dr. Koa Eher 的两分钟论文解读。
Koa Eher. The paper is 78 pages. It is a lovely read. Perfect bedtime reading. if you don't intend to sleep. And I am not an expert. I am just a student. I could be wrong, but I am trying my best. So, let's have a look. In one test, Alpha Genome 2 could identify double the number of mutations known to affect gene activity with 90% precision. For every 10 mutations the AI marked, about nine were likely to affect gene activity.
Koa Eher。这篇论文有 78 页。这是一篇令人愉悦的阅读材料。完美的睡前读物,如果你不打算睡觉的话。我不是专家,我只是一个学生。我可能会出错,但我正在尽力。所以,让我们来看看。在一项测试中,Alpha Genome 2 能够以 90% 的精度识别出已知影响基因活性的突变数量翻倍。对于 AI 标记的每 10 个突变,大约有 9 个很可能影响基因活性。
That is incredible. Now, question two. Does it generalize? Now, hold on to your papers, fellow scholars, because the answer is yes. In a Parkinson's disease study, a previous specialist AI could correctly predict whether gene activity went up or down for 17% of the known genetic effects. And alpha genome 2 had never trained on that study. So, how many could it predict? What are you seeing? What I am seeing? 41% more than double both at the same high level of accuracy.
这太不可思议了。现在,问题二:它能泛化吗?各位学者请坐稳了,因为答案是肯定的。在一项帕金森病研究中,之前的专家级 AI 只能正确预测 17% 的已知遗传效应中基因活性是上升还是下降。而 Alpha Genome 2 从未在该研究上进行过训练。那么它能预测多少呢?你看到了什么?我看到了什么?41%,比之前高出两倍,同时保持了同样高的准确率。
So this AI performed better on completely new information than a specialist AI that was trained on this information. Folks, this is science fiction level of goodness. Absolutely insane work. Okay, but we do the real talk here. Limitations. It is far from perfect. Even its best predictions still miss many disease related mutations and it's still not great at rare cell types. So please do not walk away from this video thinking that they've solved genetic diseases.
因此,这项 AI 在完全新的信息上表现优于针对该信息训练的专家级 AI。朋友们,这是科幻级别的优秀成果。绝对疯狂的工作。好吧,但我们这里要说点实在的。局限性。它还远非完美。即使它最好的预测仍然遗漏了许多与疾病相关的突变,并且它在稀有细胞类型上的表现仍然不佳。所以请不要看完这个视频后认为他们已经解决了遗传性疾病。
Not even close. But when AlphaFold first came out, I remember thinking that it had all the makings of a Nobel Prize. And looking at this new work, I wouldn't be surprised if DeepMind wins another. I would like to congratulate Sir Demi Hassabis and Pushid Kohi and the whole team. Sometimes the difference between healthy and sick is just a single letter and understanding that letter can change millions and millions of lives and I [clears throat] hope that this work will save us from terrible diseases in the future and that is the best application of AI ever.
差得远呢。但当我第一次看到 AlphaFold 发布时,我记得心想它具备了获得诺贝尔奖的所有潜质。看着这项新工作,如果 DeepMind 再获一项殊荣,我也不会感到惊讶。我想祝贺 Demi Hassabis 爵士、Pushmeet Kohli 以及整个团队。有时健康与疾病的区别仅仅是一个字母,理解那个字母可以改变数百万人的生活,我[清嗓子]希望这项工作将来能帮助我们摆脱可怕的疾病,这是 AI 有史以来最好的应用。
Oh, and the best part, Alpha Genome 2 is already available to researchers for free through an API and DeepMind plans to release the code and model weights too. What a time to be alive. And I have a recent interview with the creator Pushmeat talking about alpha genome 1 in the video description if you are interested and subscribe for more. I use Lambda to reproduce AI research papers often in minutes. It's great to train your own models, fine-tune an existing one, run inference, text to image, or video.
哦,还有最棒的部分:Alpha Genome 2 已经通过 API 向研究人员免费提供,DeepMind 还计划发布代码和模型权重。生逢此时,何其有幸。如果你对 Alpha Genome 1 感兴趣,视频简介中有我最近对创造者 Pushmeat 的采访,欢迎订阅以获取更多内容。我经常使用 Lambda 在几分钟内复现 AI 研究论文。训练自己的模型、微调现有模型、运行推理、文本生成图像或视频,都非常出色。
Easy peasy. Running a Deep Seek chatbot or agent. Super fast, super reliable. Lambda gives you powerful Nvidia GPUs to run your own experiments. I test ideas from papers I cover and moments later, results. Love it. Seriously, try it out now at lambda.com/papers.
轻松搞定。运行 Deep Seek 聊天机器人或智能体。速度超快,可靠性极高。Lambda 为你提供强大的 Nvidia GPU,用于运行你自己的实验。我会测试我所报道论文中的想法,片刻之后就能看到结果。非常喜欢。说真的,现在就去 lambda.com/papers 试试看吧。
更进一步:量化金融体系
看懂新闻只是起点——沿量化金融路径,把它变成能交付的工程能力