These unfolded protein structures then start interacting with other clusters of proteins, called caseins.

进而使得这些展开的蛋白结构开始与其他蛋白质集群相互作用产生酪蛋白。

Turns out, the bigger and more numerous the pores within the blocks' protein structure the more salt that was released.

结果他们发现，块状蛋白结构气孔越大越多，释放的盐分也就越多。

Many common colds are caused by rhinoviruses that have a geometric protein structure called a capsid instead of a lipid membrane.

很多普通感冒是由鼻病毒引发的，它有一种称为“衣壳”的几何蛋白结构，而不是脂质膜结构。

Not necessarily. Many common colds are caused by rhinoviruses that have a geometric protein structure called a capsid instead of a lipid membrane.

未必。很多普通感冒是由鼻病毒引发的，它有一种称为“衣壳”的几何蛋白结构，而不是脂质膜结构。

Facebook's parent company, Meta, ditched the first database of more than 600mn protein structures.

Facebook 的母公司 Meta 放弃了第一个包含 6 亿多个蛋白质结构的数据库。

This magnifies the initial protein structure by around four-and-a-half times, while preserving its proportions.

这将初始蛋白质结构放大了大约四倍半，同时保持了其比例。

The social media giant had employed about a dozen scientists and used artificial intelligence to predict protein structures.

这家社交媒体巨头雇佣了大约十几名科学家， 并利用人工智能来预测蛋白质结构。

AlphaFold 2 now has a database of more than 200m protein structure predictions, with a prediction accuracy approaching 90%.

AlphaFold 2目前拥有超过2亿个蛋白质结构预测的数据库，预测准确率接近90%。

Most known protein structures have been worked out from crystallised versions of the molecule, held tight by networks of chemical bonds.

大多数已知的蛋白质结构都是在蛋白质分子的结晶状态下测定的，此时其结构被化学键网络固定。

Going the other way, and predicting a protein's structure from its amino-acid sequence, is a problem that took even longer to crack.

反过来说， 根据蛋白质的氨基酸序列预测其结构，则是一个需要更长时间才能解决的问题。

Since the method can be applied iteratively, he reckons it will eventually be able to create a 100-fold magnification of protein structures.

由于该方法可以反复应用，他认为它最终能够将蛋白质结构放大 100 倍。

The team plans to study crowded protein structures linked to Parkinson's disease and molecules implicated in the ageing process in the brain.

该团队计划研究与帕金森病相关的拥挤蛋白质结构以及与大脑衰老过程有关的分子。

And this year, researchers at DeepMind solved 350,000 human protein structures — 44% of all known human proteins, and a really good fraction of the 200 million thought to exist.

今年，DeepMind 的研究人员破解了 35 万个人类蛋白质结构——占所有已知人类蛋白质结构的 44%，是被认为存在的 2 亿种蛋白质中的相当多一部分。

As it turns out, this specialized insulin could help treat Type I diabetes someday, and it’s all thanks to some fancy folding in the protein structure.

事实证明，这种专门的胰岛素有朝一日可以帮助治疗 I 型糖尿病， 这全都归功于蛋白质结构中的一些奇特折叠。

Nonetheless—and depending on how DeepMind decides to license the technology—an ability to generate protein structures routinely in this way could have a big impact on the field.

尽管如此——并且取决于 DeepMind 决定如何授权该技术——以这种方式常规生成蛋白质结构的能力可能对该领域产生重大影响。

Now, the game—researchers developed Foldit in 2008. Teams of players fold molecules and rotate amino acids to create 3-D protein structures. Gamers get points for structure stability.

现在，我们说明这档2008年推出的游戏——佛德特（Foldit，学名：实验性的蛋白质折叠电子游戏），玩家可以各自分小组进行，所有玩家在游戏中折叠分子和旋转氨基酸，创造出三维的蛋白结构，游戏得分标准基于蛋白质的稳定程度。

And similarly, Pavlov's dogs probably don't know a lot about, like, protein structures or the waveform of a ringing bell, but they salivate nonetheless because the body paired the stimuli.

同样，巴甫洛夫的狗可能不太了解蛋白质结构或铃声的波形，但它们仍然会分泌唾液，因为身体会与刺激配对。

Scientists wanted to know the structure of a protein-cutting enzyme produced by a retrovirus similar to HIV. They haven't been able to solve the protein structure using standard computational methods.

科学家们很想掌握由反向病毒所衍生的蛋白剪切酶的结构，而反向病毒是指，比如HIV这样的病毒。这类蛋白结构是不能通过标准算法法则推导就直接得出的。

Demis Hassabis and John Jumper from DeepMind, Google's AI company, received one half of the chemistry prize for their development of AlphaFold, a program capable of predicting three-dimensional protein structure, a long-standing grand challenge in biochemistry.

谷歌人工智能公司 DeepMind 的 Demis Hassabis 和 John Jumper 因开发 AlphaFold 而获得了一半化学奖。AlphaFold 是一个能够预测三维蛋白质结构的程序，这是生物化学领域长期以来的一大挑战。