Add BrightSurf on Google Email

AI as a tool in flavor research

Researchers developed an AI-based method to predict bitterness of peptides, enabling de novo design of bitter peptides for improved flavor control. The method used a combination of a protein language model and an artificial neural network to analyze structural data, resulting in the identification of new bitter-tasting peptides.

SourceLeibniz-Institut für Lebensmittel-Systembiologie an der TU München·Journalnpj Science of Food·TypeComputational simulation/modeling·DateSep 22, 2026

Improving protein folding may lead to new diabetes treatments

Researchers at Sanford Burnham Prebys Medical Discovery Institute found that enhancing protein-folding processes can help prevent harm to insulin-producing cells. The study revealed the importance of a chaperone protein called binding immunoglobulin protein and its cochaperone protein p58 IPK in maintaining proper proinsulin folding.

SourceSanford Burnham Prebys·JournalProceedings of the National Academy of Sciences·TypeExperimental study·DateJun 11, 2026

Order from disordered proteins

A team of researchers developed a computational method that can design intrinsically disordered proteins with desired properties. The work uses automatic differentiation to optimize protein sequences and leverages molecular dynamics simulations for precision. This breakthrough has the potential to reveal new insights into diseases like...

SourceHarvard John A. Paulson School of Engineering and Applied Sciences·JournalNature Computational Science·TypeComputational simulation/modeling·DateOct 6, 2025

Unfolded Protein Response pathway offers new targets to treat bone weakness in cancer patients

Researchers identify three primary UPR pathways and their downstream cascades, which play a crucial role in the differentiation of osteoblasts and osteoclasts. Targeting these pathways with emerging drugs may alleviate bone-related events and kill tumors localized in bones.

SourceEditorial Office of West China School of Stomatology, Sichuan University·JournalBone Research·TypeLiterature review·DateSep 25, 2025

New class of protein misfolding simulated in high definition

Researchers at Penn State have simulated a new class of protein misfolding using atomic-scale models, revealing a type of entanglement that disrupts protein function and persists in cells. The findings support the existence of this long-lasting type of misfolding, which is thought to contribute to aging and disease.

SourcePenn State·JournalScience Advances·TypeComputational simulation/modeling·DateAug 8, 2025

Membrane anchor suppresses protein aggregation

Researchers have developed new models to explore the role of a membrane anchor on the folding and aggregation of PrP. Anchoring stabilizes folding and inhibits aggregation, with clumping induced by pre-formed aggregates, suggesting a potential mechanism for infectious prion diseases.

SourceRuhr-University Bochum·JournalProceedings of the National Academy of Sciences·TypeExperimental study·DateJan 6, 2025

Researchers discover how enzymes ‘tie the knot’

Scientists used artificial intelligence and molecular dynamics simulations to understand how enzymes fold lasso peptides into a unique structure. They identified key residues important for interaction with the substrate, enabling the design of new cyclase variants that can produce potent lasso peptides.

SourceCarl R. Woese Institute for Genomic Biology, University of Illinois at Urbana-Champaign·JournalNature Chemical Biology·TypeExperimental study·DateSep 20, 2024

Protein mutant stability can be inferred from AI-predicted structures

Researchers used AlphaFold2 to predict structural effects of mutations on protein stability, finding correlations between small structural changes and stability changes. This breakthrough opens up new possibilities for protein engineering, enabling scientists to design proteins with specific functions more effectively.

SourceInstitute for Basic Science·JournalPhysical Review Letters·TypeComputational simulation/modeling·DateAug 28, 2024

Preventing cancer cells from colonizing the liver

Cancer cells can attach themselves to liver cells when specific proteins are present, allowing them to colonize and form new tumors. This discovery provides insights into the metastatic process and may lead to potential treatments that prevent cancer from establishing new tumors.

SourceETH Zurich·JournalNature·TypeExperimental study·DateJul 24, 2024

U of T researchers develop deep-learning model that outperforms Google AI system to predict peptide structures

Researchers at U of T have developed a deep-learning model called PepFlow that can predict the full range of conformations for peptides, which are shorter than proteins but perform similar biological functions. The model combines machine learning and physics to capture precise and accurate conformations within minutes.

SourceUniversity of Toronto·JournalNature Machine Intelligence·DateJun 27, 2024

By listening, scientists learn how a protein folds

Researchers use data sonification to convert molecular data into sounds, revealing how hydrogen bonds contribute to protein folding. The process involves complex interactions between water molecules and amino acids, with faster bonds speeding up folding and slower ones slowing it down.

SourceUniversity of Illinois at Urbana-Champaign, News Bureau·JournalProceedings of the National Academy of Sciences·TypeComputational simulation/modeling·DateMay 20, 2024

Rice study reveals insights into protein evolution

Researchers at Rice University have made a groundbreaking discovery about protein evolution, revealing that pseudogenes can provide clues to the evolutionary journey of proteins. The team found that certain mutations can stabilize the folding of pseudogenes, but also disrupt their biological functions.

SourceRice University·JournalProceedings of the National Academy of Sciences·DateMay 13, 2024

Fast folding for synthetic peptides and microproteins

Researchers at Xi'an Jiaotong-Liverpool University developed a new method that enables the efficient production of cysteine-rich peptides and microproteins in their naturally folded 3D structure. The approach uses organic solvents to mimic nature's oxidative folding process, resulting in speeds of over 100,000 times faster than aqueous...

SourceXi'an Jiaotong-Liverpool University·JournalAngewandte Chemie·TypeExperimental study·DateMar 21, 2024