Designing protein sequences of both high fitness and novelty is a challenging task in data-efficient protein engineering. Exploration beyond wild-type neighborhoods often leads to biologically implausible sequences or relies on surrogate models that lose fidelity in novel regions. Here, we propose PROSPERO, an active learning framework in which a frozen pre-trained generative model is guided by a surrogate updated from oracle feedback. By integrating fitness-relevant residue selection with biologically-constrained Sequential Monte Carlo sampling, our approach enables exploration beyond wild-type neighborhoods while preserving biological plausibility. We show that our framework remains effective even when the surrogate is misspecified. PROSPERO consistently outperforms or matches existing methods across diverse protein engineering tasks, retrieving sequences of both high fitness and novelty.

Designing protein sequences of both high fitness and novelty is a challenging task in data-efficient protein engineering. Exploration beyond wild-type neighborhoods often leads to biologically implausible sequences or relies on surrogate models that lose fidelity in novel regions. Here, we propose ProSpero, an active learning framework in which a frozen pre-trained generative model is guided by a surrogate updated from oracle feedback. By integrating fitness-relevant residue selection with biologically-constrained Sequential Monte Carlo sampling, our approach enables exploration beyond wild-type neighborhoods while preserving biological plausibility. We show that our framework remains effective even when the surrogate is misspecified. ProSpero consistently outperforms or matches existing methods across diverse protein engineering tasks, retrieving sequences of both high fitness and novelty.

Artificial intelligence is rapidly conquering agriculture and the food industry. To feed billions of people, you need a lot of lands. It is impossible to cultivate it manually these days. At the same time, plant diseases and insect invasions often lead to crop failures. With the modern scale of agriculture business, such invasions are difficult to identify and neutralize in the bud in time.

We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Organizations are increasingly adopting AI-enabled technologies to address existing and emerging problems within the enterprise ecosystem, meet changing market demands and deliver business outcomes at scale. Shubhangi Vashisth, senior principal research analyst at Gartner, said that AI innovation is happening at a rapid pace. Vashisth further noted that innovations including edge AI, computer vision, decision intelligence and machine learning will have a transformational impact on the market in coming years. However, while AI-powered technologies are helping to build more agile and effective enterprise systems, they usher in new challenges. For example, Gartner notes that AI-based approaches if left unchecked can perpetuate bias, leading to issues, loss of productivity and revenue.

Science fiction is a genre very much associated with technological marvels, innovations, and visions of the future. So it may be surprising to find so many of its writers are drawn to Shakespeare – he's a figure associated with tradition and the past. Sometimes his plays are reworked in a science fiction setting. The 1956 film Forbidden Planet is just one of many variations on a "Tempest in space" theme. Sometimes the playwright appears as a character caught up in a time travel adventure.

When their inventor the aptly named Dr Capek dies, the robot who plays Hamlet determines to find out the truth and just like Shakespeares original prince avenge the murder of his creator.This is just one example of a strange apparent association between Hamlet and robots.

Science fiction is a genre very much associated with technological marvels, innovations, and visions of the future. So it may be surprising to find so many of its writers are drawn to Shakespeare – he's a figure associated with tradition and the past. Sometimes his plays are reworked in a science fiction setting. The 1956 film Forbidden Planet is just one of many variations on a "Tempest in space" theme. Sometimes the playwright appears as a character caught up in a time travel adventure.

Science fiction is a genre very much associated with technological marvels, innovations, and visions of the future. So it may be surprising to find so many of its writers are drawn to Shakespeare – he's a figure associated with tradition and the past. Sometimes his plays are reworked in a science fiction setting. The 1956 film Forbidden Planet is just one of many variations on a "Tempest in space" theme. Sometimes the playwright appears as a character caught up in a time travel adventure.

In this special issue we are reprinting our top stories of the past year. This article first appeared online in our "Hidden Truths" issue in June, 2021. If you could soar high in the sky, as red kites often do in search of prey, and look down at the domain of all things known and yet to be known, you would see something very curious: a vast class of things that science has so far almost entirely neglected. These things are central to our understanding of physical reality, both at the everyday level and at the level of the most fundamental phenomena in physics--yet they have traditionally been regarded as impossible to incorporate into fundamental scientific explanations. They are facts not about what is--"the actual"--but about what could or could not be. In order to distinguish them from the actual, they are called counterfactuals. Suppose that some future space mission visited a remote planet in another solar system, and that they left a stainless-steel box there, containing among other things the critical edition of, say, William Blake's poems.

If you could soar high in the sky, as red kites often do in search of prey, and look down at the domain of all things known and yet to be known, you would see something very curious: a vast class of things that science has so far almost entirely neglected. These things are central to our understanding of physical reality, both at the everyday level and at the level of the most fundamental phenomena in physics--yet they have traditionally been regarded as impossible to incorporate into fundamental scientific explanations. They are facts not about what is--"the actual"--but about what could or could not be. In order to distinguish them from the actual, they are called counterfactuals. Suppose that some future space mission visited a remote planet in another solar system, and that they left a stainless-steel box there, containing among other things the critical edition of, say, William Blake's poems. That the poetry book is subsequently sitting somewhere on that planet is a factual property of it. That the words in it could be read is a counterfactual property, which is true regardless of whether those words will ever be read by anyone.