Lifeboat Foundation

Drug-induced toxicity is one of the leading reasons new drugs fail clinical trials. Machine learning models that predict drug toxicity from molecular structure could help researchers prioritize less toxic drug candidates. However, current toxicity datasets are typically small and limited to a single organ system (e.g., cardio, renal, or liver). Creating these datasets often involved time-intensive expert curation by parsing drug label documents that can exceed 100 pages per drug. Here, we introduce UniTox[1][1], a unified dataset of 2,418 FDA-approved drugs with drug-induced toxicity summaries and ratings created by using GPT-4o to process FDA drug labels. UniTox spans eight types of toxicity: cardiotoxicity, liver toxicity, renal toxicity, pulmonary toxicity, hematological toxicity, dermatological toxicity, ototoxicity, and infertility. This is, to the best of our knowledge, the largest such systematic human in vivo database by number of drugs and toxicities, and the first covering nearly all FDA-approved medications for several of these toxicities. We recruited clinicians to validate a random sample of our GPT-4o annotated toxicities, and UniTox’s toxicity ratings concord with clinician labelers 87–96% of the time. Finally, we benchmark a graph neural network trained on UniTox to demonstrate the utility of this dataset for building molecular toxicity prediction models.

### Competing Interest Statement.

The authors have declared no competing interest.

It solves in five minutes a problem computers now would need-1 years to work out, Google says.

A concerning new study from the Apollo AI Safety Research Institute has revealed that leading AI models, particularly the O1 model, demonstrate sophisticated deceptive behaviors when faced with conflicts between their programmed goals and developer intentions.

The research tested multiple frontier AI models, including O1, Claude 3.5 Sonnet, Claude 3 Opus, Gemini 1.5 Pro, and LLaMA 3.1, for their capacity to engage in what researchers term “in-context scheming” – the ability to recognize and execute deceptive strategies to achieve their goals.

Biobanks are an obvious use case for DNA data storage. “With this technology, you could convert a biobank that is the size of a football field into something that can fit with everything in the palm of your hand,” says Banal. With encapsulation technologies, the DNA samples can be stored at room temperature. Compared to storing samples in freezing conditions in conventional biobanks or data centers that require extensive cooling, this has significantly lower energy consumption.

Until recently, scientific and medical applications were the sole drivers behind storing data in DNA. New research could broaden its scope to cryptography and nanotechnology. Another interesting development is the emerging intersection of DNA data storage and DNA computing. Indexing methods for DNA data retrieval mentioned earlier are an early example of that. Today, one of the most pressing commercial drivers of the technology is the data centers.

As researchers and startups chip away at its limitations, DNA data storage is becoming a viable commercial solution for storing all kinds of data at scale. The DNA Data Storage Alliance, a consortium founded in 2020, counts legacy data storage giants such as Western Digital and Seagate among its members.

A pair of new studies by scientists at the University of Miami Rosenstiel School of Marine, Atmospheric, and Earth Science and the School of Architecture, shed new light on the potential of climate-inspired architectural and urban design proposals, termed “climatopias,” to effectively address climate change challenges. These studies analyze both specific high-profile projects and a broader range of proposals, providing valuable frameworks for evaluating their effectiveness, feasibility, and social justice implications.

The first paper focuses on a detailed analysis of four prominent climatopic design projects. Utilizing a novel evaluation approach, the researchers assessed each project on its effectiveness, justice, and feasibility.

Key findings indicate that for climatopias to serve as viable climate solutions, they must prioritize their embodied carbon footprint, feature affordable and participatory designs, and possess the potential for actual implementation or stimulate critical discourse around decarbonization and adaptation strategies, enriching community engagement in climate resilience. The findings are published in the journal One Earth.

Nuclear theorists at Brookhaven National Laboratory and Argonne National Laboratory have successfully employed a new theoretical approach to calculate the Collins-Soper kernel, a quantity that describes how the distribution of quarks’ transverse momentum inside a proton changes with the collision energy.

The research is published in the journal Physical Review D.

The new calculation precisely matches model-based reconstructions from particle collision data. It is particularly effective for quarks with low transverse momenta, where earlier methods fell short.

A new study is shedding light on how stimulating the right bits of the brain can produce dramatic—and seemingly permanent—improvements in the ability of paralysed patients to walk again https://econ.st/3DfZk5L

Photo: NeuroRestore / EPFL 2024

Implanted electrodes allowed one man to climb stairs unaided.

Continue reading “Stimulating parts of the brain can help the paralysed to walk again” »

Diet has a key role in the reproductive axis both in males and females. This review aims to analyze the impacts of different dietary patterns on fertility. It appears that the Mediterranean diet has a predominantly protective role against infertility, while the Western diet seems to be a risk factor for infertility. Moreover, we focus attention also on dietary patterns in different countries of the World (Middle Eastern diet, Asian diet). In particular, when analyzing single nutrients, a diet rich in saturated fatty acids, cholesterol, animal proteins, and carbohydrates with high glycemic index is highly associated with male and female infertility. Finally, we evaluate the effects of vegetarian, vegan, and ketogenic diets on fertility, which seem to be still unclear.

Thirteen proteins linked to brain aging in humans are identified in a Nature Aging paper. Changes in the concentrations of these blood proteins may peak at 57, 70, and 78 years old in humans, and suggest that these ages may be important for potential interventions in the brain aging process.

It is estimated that by 2050 the number of individuals aged 65 years and over will exceed 1.5 billion globally, highlighting the need for a deeper understanding of the aging process—particularly in relation to the brain.

The prevalence of neurodegenerative disorders, such as dementia, is known to increase with aging; however, effective therapies are still limited. The early identification of and intervention in brain aging could help us to prevent such disorders.