Lifeboat Foundation

Talk to any structural biologist, and they’ll tell you how a cool new method is taking over their field. By flash freezing proteins and bombarding them with electrons, cryo–electron microscopy (cryo-EM) can map protein shapes with near-atomic resolution, offering clues to their function and revealing bumps and valleys that drug developers can target. The technique can catch wriggly proteins in multiple configurations, and it can even capture those that have been off-limits to traditional x-ray analysis because they stubbornly resist being crystallized. Many researchers expect cryo-EM will surpass x-ray crystallography in the number of new protein structures solved next year.

Yet for all its charms, cryo-EM has flaws: The freezing process is finicky, and the microscopes are expensive. High-end machines can cost more than $5 million to buy, about as much to install, and hundreds of thousands per year to operate and maintain. Many U.S. states—and countries—don’t have a single cryo-EM microscope. “The haves and have-nots is what it is right now,” says Rakhi Rajan, a structural biologist at the University of Oklahoma, which currently lacks one.

Researchers at the Medical Research Council’s Laboratory of Molecular Biology (LMB) have been working to democratize the field. Today, in the, the U.K. team describes cobbling together a prototype cryo-EM microscope that has solved its first structures. The machine—what LMB physicist Chris Russo calls a “cheap little hatchback” rather than a “Ferrari”—could rival high-end machines in capabilities for one-tenth of the cost.

A method that provides cloned embryos with a healthy placenta could pave the way for more research involving the primates.

Innovative! Drought resistant and doesn’t need a lot of fertilizer. And it smells like bread. Especially good for Africa which was mentioned.

This fruit curiously smells like baking bread and tastes like potatoes – and it might also provide a means to help feed communities amid drought and famine.

Breadfruit, despite its name, is not made of bread. It’s a tall tree native to islands in the Pacific that produces a starchy fruit, similar to a jackfruit.

Continue reading “Scientists think this weird-looking, odd-smelling fruit holds the key to fighting world hunger as temperatures rise” »

The OpenDAC project is a collaborative research project between Fundamental AI Research (FAIR) at Meta and Georgia Tech, aimed at significantly reducing the cost of Direct Air Capture (DAC).

Direct Air Capture (DAC) involves directly capturing carbon dioxide from the atmosphere and has been widely recognized as a crucial tool in combating climate change. Despite its potential, the broad implementation of DAC has been impeded by high capture costs. Central to overcoming this hurdle is the discovery of novel sorbents — materials that pull carbon dioxide from the air. Discovering new sorbents holds the key to reducing capture costs and scaling DAC to meaningfully impact global carbon emissions.

The DAC space is growing rapidly with many companies entering the space. To engage the broader research community as well as the budding DAC industry, we have released the OpenDAC 2023 (ODAC23) dataset to train ML models. ODAC23 contains nearly 40M DFT calculations from 170K DFT relaxations involving Metal Organic Frameworks (MOFs) with carbon dioxide and water adsorbates. We have also released baseline ML models trained on this dataset.

Large language models (LLMs), deep learning-based models trained to generate, summarize, translate and process written texts, have gained significant attention after the release of Open AI’s conversational platform ChatGPT. While ChatGPT and similar platforms are now widely used for a wide range of applications, they could be vulnerable to a specific type of cyberattack producing biased, unreliable or even offensive responses.

Researchers at Hong Kong University of Science and Technology, University of Science and Technology of China, Tsinghua University and Microsoft Research Asia recently carried out a study investigating the potential impact of these attacks and techniques that could protect models against them. Their paper, published in Nature Machine Intelligence, introduces a new psychology-inspired technique that could help to protect ChatGPT and similar LLM-based conversational platforms from cyberattacks.

“ChatGPT is a societally impactful artificial intelligence tool with millions of users and integration into products such as Bing,” Yueqi Xie, Jingwei Yi and their colleagues write in their paper. “However, the emergence of jailbreak attacks notably threatens its responsible and secure use. Jailbreak attacks use adversarial prompts to bypass ChatGPT’s ethics safeguards and engender harmful responses.”

A Northwestern University team has demonstrated a remarkable new way to generate electricity, with a paperback-sized device that nestles in soil and harvests power created as microbes break down dirt – for as long as there’s carbon in the soil.

Microbial fuel cells, as they’re called, have been around for more than 100 years. They work a little like a battery, with an anode, cathode and electrolyte – but rather than drawing electricity from chemical sources, they work with bacteria that naturally donate electrons to nearby conductors as they chow down on soil.

The issue thus far has been keeping them supplied with water and oxygen, while being buried in the dirt. “Although MFCs have existed as a concept for more than a century, their unreliable performance and low output power have stymied efforts to make practical use of them, especially in low-moisture conditions,” said UNW alumnus and project lead Bill Yen.

A group of over 400 scientists have spent the last decade studying supernovae with unprecedented results about the expansion of space and the role of dark energy.

The young sun may have captured several Mars-or Mercury-size exoplanets that now orbit in the outer reaches of the solar system, but identifying them will be extremely challenging.

Backpropagation is the key algorithm that makes training deep models computationally tractable. For modern neural networks, it can make training with gradient descent as much as ten million times faster, relative to a naive implementation. That’s the difference between a model taking a week to train and taking 200,000 years.

Beyond its use in deep learning, backpropagation is a powerful computational tool in many other areas, ranging from weather forecasting to analyzing numerical stability – it just goes by different names. In fact, the algorithm has been reinvented at least dozens of times in different fields (see Griewank (2010)). The general, application independent, name is “reverse-mode differentiation.”

Fundamentally, it’s a technique for calculating derivatives quickly. And it’s an essential trick to have in your bag, not only in deep learning, but in a wide variety of numerical computing situations.