swagedk

Find patients with SCA48 cased by STUB1 mutations.

[David R. Liu @davidrliu]

Base editing and prime editing clinical trial landscape, including today’s GSD type I clinical readout from @BeamTx.

[Andrej Karpathy @karpathy]

We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:

Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:

Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:

Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:

Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!

In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.

[David R. Liu @davidrliu]

First announced patient dosing of an in vivo prime editing treatment (to correct ATP7B H1069Q, one of the most common genetic causes of Wilson disease). There are now at least three prime editing clinical trials underway—for CGD (ex vivo, with positive clinical outcomes), for Wilson disease (in vivo), and for alpha-1 antitrypsin deficiency (AATD, in vivo).

[火山哥🕊️ @huoshan007]

兄弟们爬虫界要翻天了，这个太牛逼了。

browser-use 我挖出来的时候还没这么火，刚看了眼，已经 9.5 万星了。

两位苏黎世 ETH 的研究员搞的，MIT 开源。

说白了就是让 AI 像真人一样操控浏览器：

点击、滚动、登录、填表、换页面、从没见过的网站里抓数据。

我自己试过以后才敢说，这东西对普通爬虫挺狠。

很多传统爬虫摸不到的页面，尤其是动态渲染、要交互、要登录的地方，它直接让 AI 上手点。

最关键的不是“能爬”，而是它开始会判断页面该怎么走。

这点有点吓人。

链接：

https://t.co/S7ELiYf4SO

[Federico Viticci @viticci]

I gave Astra a portrait of me.

And I watched as it used Apple Notes on my Mac to draw me.

This model feels incredible.

[馬修 @idoubaody]

算力成本暴跌 95% ！ 菲尔兹奖得主把大模型做成“心灵感应”，硅谷巨头被指着鼻子打脸。

连线杂志消息，几个顶级俄罗斯数学大佬（甚至搬出了数学界诺贝尔奖“菲尔兹奖”得主），搞出了一套让 AI 之间“心灵感应”的黑科技，硬生生把一个手机就能跑的 40 亿模型和一个 7500 亿参数的超级巨无霸直接“合体”——算力成本瞬间暴跌 95%，性能却一点不拉胯，甚至还顺手把全球最难的 AI 推理大赛（ARC-AGI）给刷到了榜首！

最可怕的是，它直接指着硅谷那些天天烧几百亿堆算力、搞封锁的巨头鼻子打脸：以后根本不需要再无限膨胀单体大模型，靠“小模型白嫖大模型脑子”就能逆袭！

这不仅把 AI 商业化的最大痛点（太贵）给一枪干爆，更是给开源阵营递了一把反杀闭源巨头的超级武器，彻底把全行业的玩法给颠覆了！

[David R. Liu @davidrliu]

Today in @NatureBiotech we report the use of phage-assisted evolution (PACE) to reprogram botulinum neurotoxin (BoNT) proteases to trigger cancer cell death. We evolved proteases to cleave procaspase-1 or gasdermin D, inducing cell death and slowing tumor growth in a highly drug-resistant mouse model.
https://t.co/OTUVxl1FpG
1/13

[David R. Liu @davidrliu]

Amazing new work by @JShendure and coworkers uses a clever application of prime editing called DNA Typewriter to trace the lineage of a developing mouse from zygote to middle of day 14 at a single-cell level. See Jay’s beautiful tweetorial:

[David R. Liu @davidrliu]

Today in @Nature we report how AI-guided redesign enhances protein evolution. Integrating ProteinMPNN sequence design with autonomous laboratory evolution, we establish a workflow to engineer enzymes with improved properties over those evolved from natural proteins. Redesigned starting points consistently evolve an expanded fitness landscape, reaching new function with higher activity, specificity, and stability than their natural counterparts.
https://t.co/pNlw6z1zvU
1/14

[路飞 🏴‍☠️ AI 研究员🧐 @0xluffy_eth]

一位前Anthropic研究员泄露了团队使用的内部提示框架。

大多数人把Claude当基础聊天机器人，浪费了60-70%的推理能力。

这10个提示是专业人士的实际用法——内部测试过，最大化清晰度、诚实度和深度。

复制即用。零废话。

保存此线程。你的Claude水平即将永久改变。

(专业提示：按顺序使用以获得复合效果)

[David R. Liu @davidrliu]

Today in @NatureBiotech we report a new suit of PE8 prime editor proteins. PE8 variants were developed from laboratory-evolved PE6 proteins using AI-guided protein redesign. This approach combines recent advances in computational protein design and directed evolution to increase prime editing efficiency, especially in transient therapeutically relevant delivery settings such as mRNA+pegRNA electroporation into primary cells, eVLP delivery of prime editing RNPs, and LNP-mediated mRNA+pegRNA delivery in mice.
https://t.co/bz6PalFvc4
1/11

[David R. Liu @davidrliu]

Today in @NatBiotech, we report the directed evolution of structured RNA motifs that enhance the efficiency of prime editing. Iterated high-throughput pooled screens and mutagenesis of these small RNA elements improved transient pegRNA lifetime.
https://t.co/ZiytYXicaJ
1/11

[David R. Liu @davidrliu]

I’m honored to have been presented the Franklin Institute Bower Award for Achievement in Science by Kyle and Nicole Muldoon, the parents of baby K.J. (Kyle Junior), alongside @kiranmusunuru and @AhrensNicklas. Video from the ceremony:
https://t.co/fPud2WTgvY

[David R. Liu @davidrliu]

Today in @natBME we report an in vivo base editing strategy that corrects a common disease-causing mutation and rescues pathology in a mouse model of Zellweger spectrum disorder (ZSD) and restore peroxisomal function in patient derived cells. This work highlights the promise of precision genome editing for treating peroxisomal biogenesis disorders.
1/13

[David R. Liu @davidrliu]

I'm honored to be included in the @washingtonpost Next 50, representing a remarkable community—scientists, physicians, caregivers, and the patients who inspire us—working together to bring the life-saving promise of therapeutic gene editing to those living with genetic diseases.

Profile by Carolyn Johnson:
https://t.co/bJ6nqTFWP1

https://t.co/1ZQEbKUIiF

[David R. Liu @davidrliu]

Below is the story of the first patient treated with a prime-edited therapeutic, developed by @PrimeMedicine in a trial led by Dr. Élie Haddad and his team at CHU Sainte-Justine. This teenager suffered from chronic granulomatous disease (CGD), an immunodeficiency, and now—10 months after treatment—the patient is healthy, stable, and living with a functioning immune system. Tracy Attebury, whose story was previously told by @ginakolata @nytimes, was the second patient treated with a prime-edited therapeutic.

https://t.co/ZiVq1SmYsn

[David R. Liu @davidrliu]

From @NatureBiotech: New rules spur cell and gene therapy trials in China

“Once proven safe and effective in investigator-initiated trials, therapies [in China] can then be given to patients as treatments, again without requiring national regulatory approval. And the hospital that made the therapy and conducted the trial can charge for the treatment, creating a financial incentive for hospitals to deploy high-quality cell and gene therapies.

This situation contrasts with that in individual European Union countries, for example, where regulations enable hospitals to deliver investigational cell therapies when few or no approved alternatives remain. But, unlike in China, national regulatory approval is usually needed to go ahead with testing…

In the United States, the process is more onerous. Hospital-developed cell therapies destined for patients require an Investigational New Drug application approved by the US Food and Drug Administration, with some exceptions for compassionate use.”

@US_FDA @HHSGov

https://t.co/UVLks7nncL

[The Eagle flies free @Fa21519230]

🚨🚨 La CIA tradujo en secreto un artículo soviético: los parásitos y el cáncer se comportan bioquímicamente de forma casi idéntica...Los mismos antiparasitarios (Myracil D, análogos de la guanina) eliminaron a la vez la esquistosomiasis y los tumores en el laboratorio...
Esto se clasificó como confidencial y se guardó durante 60 años en una caja fuerte...
Investigaciones actuales exploran el uso de antiparasitarios como ivermectina y fenbendazol para combatir el cáncer, a raíz de las similitudes metabólicas observadas en el artículo de 1951...

[🌹MIDNIGHT🌹ROSE🌹 @369Jessica369]

🚨一位日本科学家因发现人体最强大的自愈机制之一而荣获诺贝尔奖：在禁食期间，人体会开始吞噬自身受损细胞，从而触发一种被称为自噬的深度细胞重置过程。这一自然过程如同一个内部清洁系统，能够分解衰老和疾病相关的衰老蛋白质、功能失调的细胞成分以及毒素积累。这项发现表明，禁食不仅仅改变新陈代谢，它还能激活我们生物体中固有的基本生存程序。

研究人员发现，自噬能够促进细胞更新、支持免疫功能、预防神经退行性疾病并增强整体抵抗力。当身体不再忙于消化食物时，它会将重心转移到维护和修复上，将受损组织回收利用，转化为新鲜健康的成分。这有助于减少炎症、提升大脑功能并增强身体应对压力的能力。一些研究甚至表明，激活自噬可能降低与衰老相关的慢性疾病的风险。

这项荣获诺贝尔奖的研究彻底改变了科学家们对禁食的看法，使其从一种简单的饮食潮流转变为一种真正意义上的生物重置方式。医生们现在表示，间歇性禁食——只要安全进行——就能帮助维持长期的细胞健康，提升能量水平，并改善代谢平衡。虽然它并非万能药，但这项发现表明，在合适的条件下，人体拥有非凡的自我修复能力。

对于全世界数百万人来说，这项研究打开了一扇新的窗口，让我们得以了解我们的生活方式对细胞的深远影响，以及像安排用餐时间这样简单的事情如何能够激活人体最强大的修复系统之一。

[康客 @kangke888]

重磅进展：全球首款“治愈性”糖尿病疗法最快今年上市

仅通过一次注射治疗，就能让严重的1型糖尿病实现临床治愈？

根据国际顶级期刊《新英格兰医学杂志》近日发表的一项临床试验结果，一款名为 zimislecel 的干细胞疗法在研究中展现出惊人的疗效：参与试验的10名重症患者在接受单次注射治疗后，已彻底摆脱胰岛素注射，实现了临床意义上的功能性治愈。

这一研究成果被誉为糖尿病治疗领域的里程碑。该疗法由全球生物制药巨头Vertex Pharmaceuticals（福泰制药）研发，目前已经进入关键的III期临床试验阶段，公司透露最快有望在 2026年正式申请上市。

一次注射，终结每日胰岛素

研究数据显示，参与I/II期临床试验的14名患者均为病史悠久的重症1型糖尿病患者，平均病程超过20年。在治疗前，他们的胰岛功能几乎完全丧失，必须依赖外源性胰岛素维持生命，且饱受严重低血糖事件的困扰。

在接受全剂量zimislecel注射后：

脱离胰岛素： 10名患者在治疗后6个月至1年内完全停用胰岛素。

血糖恢复正常： 所有患者在接受治疗后4个月，糖化血红蛋白（HbA1c）水平降至正常范围。

血糖平稳： 在超过93%的监测时间内，患者的血糖水平均维持在理想范围内，有效避免了血糖剧烈波动带来的风险。

另外2名接受全剂量治疗的患者，胰岛素用量也分别大幅减少了70%和36%。研究团队指出，患者在治疗后90天内，体内C肽水平（反映胰岛功能的关键指标）逐渐恢复，表明移植的再生胰岛已经开始稳定发挥生理功能。

25年追梦：从科学家父亲到革命性疗法

zimislecel的诞生背后是一个关于爱与坚持的故事。该疗法的核心发明者、哈佛大学干细胞专家道格拉斯·梅尔顿教授投身研究的初衷，源于他6个月大的孩子被诊断出患有1型糖尿病。25年来，他带领团队历经无数次失败，最终成功将人类多能干细胞诱导分化为功能性的胰岛细胞。

从原理上看，zimislecel属于同种异体干细胞再生疗法。它利用诱导性多能干细胞技术，在体外培养出健康的胰岛β细胞，并将其移植到患者体内。这些全新的细胞会在肝脏中“安家落户”，像真正的胰岛一样实时感知血糖变化，并精准分泌胰岛素，从而实现对血糖的生理性调控。

加州大学旧金山分校糖尿病中心主任马克·安德森教授评价称：“这是一项开创性的工作，脱离胰岛素将彻底改变无数患者的生活轨迹。”

挑战与希望并存：免疫排斥与安全性考量

尽管前景光明，该疗法目前仍面临挑战。由于使用的是异体干细胞，患者在治疗后需要长期服用免疫抑制剂以防止身体对新植入的细胞产生排斥。福泰制药也承认，这一要求在临床试验招募初期曾让不少潜在参与者望而却步。

在安全性方面，研究中报告了包括腹泻、恶心等在内的轻度至中度不良事件。值得注意的是，有两位患者在糖尿病被成功治愈后因其他原因不幸去世：一位因既往低血糖事故遗留的脑损伤恶化导致痴呆离世；另一位则因无关的鼻窦炎手术引发了致命的真菌感染。