Explore

Rapid ReadRapid Read
AI-Generated Molecular Fingerprints Advance Drug Discovery for Therapeutic Compounds

AI-Generated Molecular Fingerprints Advance Drug Discovery for Therapeutic Compounds

Junmei Wang, a professor of pharmaceutical sciences in the School of Pharmacy, has developed a novel framework called Full-Spectrum Generative Lead Discovery (FSGLD) that utilizes generative AI to identify existing compounds with therapeutic potential. Published in *Science Advances* on October 28, this framework addresses the limitations of traditional AI-driven drug discovery, which often focuses on generating new molecules that are difficult to synthesize. Instead, Wang's lab employs a generative adversarial network (GAN) named DRUG-GAN to create 'molecular fingerprints'—digital encodings of a molecule's key chemical features. The GAN consists of a generator that produces fake fingerprints and a discriminator that distinguishes them from real ones. Through this training, the generator learns to create fingerprints closely resembling those of real molecules. Researchers then search chemical databases for existing compounds that match these AI-generated patterns, leading to the discovery of real, testable...
Rapid ReadRapid Read
Dead Sea Salt Composition Differs Significantly from Ordinary Sea Salt, Dominated by Magnesium and Calcium

Dead Sea Salt Composition Differs Significantly from Ordinary Sea Salt, Dominated by Magnesium and Calcium

Dead Sea salt is characterized by a unique chemical composition, differing substantially from typical sea salt. Unlike ordinary sea salt, which is predominantly sodium chloride, Dead Sea salt contains only about 30% sodium chloride. Its primary constituents are magnesium, calcium, potassium, and bromide salts. This distinct mineral profile is a result of the Dead Sea's extreme environment, a landlocked hypersaline lake in the Jordan Rift Valley. Intense heat and evaporation over extended periods have concentrated dissolved minerals, leading to this unusual composition. The Dead Sea lies approximately 440 meters below sea level, making it Earth’s lowest exposed land surface. With no outlet, water loss occurs almost entirely through evaporation, driving its salinity to about 34%, which is ten times that of the ocean. This high density, reaching around 1.24 kg per liter, allows for effortless floating and largely prevents marine life from thriving.
TrendlineTrendline
AIRCO Establishes Pennsylvania Hub for Carbon Dioxide-to-Fuel Technology

AIRCO Establishes Pennsylvania Hub for Carbon Dioxide-to-Fuel Technology

AIRCO Establishes Pennsylvania Hub for Carbon Dioxide-to-Fuel Technology
Rapid ReadRapid Read
Vanderbilt-Ingram Cancer Center Unveils Comprehensive Transcriptomic Atlas to Enhance Breast Cancer Treatment Precision

Vanderbilt-Ingram Cancer Center Unveils Comprehensive Transcriptomic Atlas to Enhance Breast Cancer Treatment Precision

Researchers at Vanderbilt-Ingram Cancer Center have developed one of the largest transcriptomic maps of breast cancer to date. This comprehensive atlas, detailed in Cell Reports Medicine, integrates 7,612 breast cancer transcriptomes from 22 independent datasets, providing an unprecedented view of the disease's molecular diversity. The team, led by Brian Lehmann, PhD, Research Associate Professor of Medicine, identified 12 breast cancer-intrinsic subtypes and three major tumor microenvironment programs, collectively named the integrated Breast Cancer Transcriptome Classification (iBTC). This atlas reveals previously unrecognized forms of breast cancer with distinct prognoses, biological characteristics, and therapeutic vulnerabilities, moving beyond traditional classifications to enable more precise and effective treatment decisions.
Rapid ReadRapid Read
Department of Energy Researchers Develop AI-Driven Robotic System to Accelerate Materials Discovery

Department of Energy Researchers Develop AI-Driven Robotic System to Accelerate Materials Discovery

Researchers at the Department of Energy have developed an advanced robotic system, named AutoBot, driven by Super Intelligence (SI) to significantly accelerate the design and discovery of new materials. This system replaces traditional, time-consuming trial-and-error methods with intelligent automation. AutoBot was used to analyze over 5,000 potential combinations for creating halide perovskite materials, which are promising for solar cells. The SI system efficiently selected the most informative experiments, requiring testing of fewer than 1% of the possibilities to accurately predict optimal manufacturing conditions. It identified humidity as the most critical factor influencing material quality, with moderate humidity levels facilitating the formation of the optimal crystal structure. This breakthrough promises to reduce the time for materials development from months or years to just days or weeks.
TrendlineTrendline
Arclin and KINETIXX Expand Partnership to Advance Kevlar® EXO™ Technology Globally

Arclin and KINETIXX Expand Partnership to Advance Kevlar® EXO™ Technology Globally

Arclin and KINETIXX Expand Partnership to Advance Kevlar® EXO™ Technology Globally
Rapid ReadRapid Read
EPA Refines Surfactant Risk Assessment with New Non-Animal Method

EPA Refines Surfactant Risk Assessment with New Non-Animal Method

The Environmental Protection Agency (EPA) has adopted a new approach for assessing the inhalation risk of surfactants, moving away from traditional animal testing. This refined method, known as a new approach method (NAM), utilizes the MucilAir™ 3D human airway epithelial model, a peer-reviewed in vitro platform. This lab-grown human airway tissue is exposed to chemicals to measure their effects on the respiratory tract, allowing the EPA to calculate human equivalent concentrations (HECs) for portal-of-entry respiratory irritation. This interim approach covers most common surfactant types, including nonionic, anionic, and zwitterionic, but currently excludes cationic surfactants. The EPA has also provided a guide for classifying substances as surfactants under this new framework. Two benchmark chemicals, Triton X-100 and Oleoyl sarcosine, have been used to derive occupational HECs, setting a conservative standard for future risk characterization.
TrendlineTrendline
Molekula Details Strategies for Isotopically Labelled Nucleotide Supply Chain Resilience Amid Geopolitical Tensions

Molekula Details Strategies for Isotopically Labelled Nucleotide Supply Chain Resilience Amid Geopolitical Tensions

Molekula Details Strategies for Isotopically Labelled Nucleotide Supply Chain Resilience Amid Geopolitical Tensions
Rapid ReadRapid Read
University of Pittsburgh Professor Develops AI Framework for Drug Discovery

University of Pittsburgh Professor Develops AI Framework for Drug Discovery

Junmei Wang, a professor of pharmaceutical sciences at the University of Pittsburgh, has developed a new generative AI framework called Full-Spectrum Generative Lead Discovery (FSGLD). This framework, detailed in a paper published in the journal Science Advances on October 28, utilizes generative AI to screen existing compounds for potential therapeutic applications. Unlike traditional AI methods that often focus on generating new, potentially unsynthesizable molecules, FSGLD identifies real, existing compounds that can bind to target receptors. The system employs a general adversarial network (GAN) called DRUG-GAN, which learns to generate molecular fingerprints. These AI-generated patterns then guide researchers in searching chemical databases for existing molecules with similar characteristics. Wang's lab successfully used this method to screen for compounds linked to G protein-coupled receptors (GPCRs) and protein kinase, two commonly targeted drug receptors, identifying novel compounds that effectivel...
Rapid ReadRapid Read
GeneSPOT Computational Framework Introduced to Map Spatial Gene Relationships and Dynamics in Tissues

GeneSPOT Computational Framework Introduced to Map Spatial Gene Relationships and Dynamics in Tissues

Researchers have developed Gene SPatial Optimal Transport, or GeneSPOT, a new computational framework designed to characterize the spatial relationships among genes and quantify their reorganization across various biological conditions and over time. Published in Nature Methods, this gene-centric analytical framework complements existing cell-centric methods by mapping how the relational spatial organization of molecular features changes within tissues. While current spatial omics technologies primarily focus on identifying cellular states and differentially expressed molecular features, GeneSPOT extends this by evaluating spatial arrangements beyond individual cellular states. The framework has successfully resolved tissue structures in diverse applications, including validation analyses in the mouse brain, where it captured biologically meaningful spatial gene programs and identified organized molecular patterns within complex tissue architecture.
Rapid ReadRapid Read
New Research Clarifies Genetic Risk in Pattern Dystrophy Using Protein Structure and AI

New Research Clarifies Genetic Risk in Pattern Dystrophy Using Protein Structure and AI

A recent medRxiv preprint, titled “Advancing genetic pathogenicity prediction with three-dimensional proteoform-phenotype analysis,” introduces a novel framework to better understand genetic risk in inherited retinal diseases, specifically pattern dystrophy. This research combines computer-generated protein structure models with clinical genetic data and expert review to identify which PRPH2 gene variants are likely harmful. The study focused on missense variants, which are DNA changes that substitute one protein building block for another, and evaluated 46 pathogenic PRPH2 missense variants against a broader pool of variants of uncertain significance (VUS). By using artificial intelligence-generated models, researchers examined where variants occurred within the protein and how they might affect its physical and biological properties. This 'human-in-the-loop' approach ensures that AI-identified patterns are biologically meaningful and consistent with real patient presentations.
Giants WireGiants Wire
What channel is Giants vs. Eagles on today? Time, TV schedule to watch Week 9 game

What channel is Giants vs. Eagles on today? Time, TV schedule to watch Week 9 game

How to watch the New York Giants vs. Philadelphia Eagles in Week 9 of the 2026 regular season on Sunday, November 8.