(Press-News.org) Key Takeaways
Researchers at Harvard and Northwestern have developed a machine learning method that can design intrinsically disordered proteins with custom properties, addressing nearly 30% of all human proteins that are currently out of reach of AI tools like AlphaFold.
The new approach uses automatic differentiation, traditionally a deep learning tool, to optimize protein sequences for desired properties.
The method opens new possibilities for engineering proteins, directly from physics-based models, that do not fold into a specific shape.
In synthetic and structural biology, advances in artificial intelligence have led to an explosion of designing new proteins with specific functions, from antibodies to blood clotting agents, by using computers to accurately predict the 3D structure of any given amino acid sequence.
But the structure of close to 30% of all proteins expressed by the human genome are challenging to predict for even the most powerful AI tools, including the Nobel-winning AlphaFold. Never settling into a fixed shape but constantly shifting around, these so-called intrinsically disordered proteins are key to countless biological functions like cross-linking molecules, sensing, or signaling, but their inherent instability makes them difficult to design from scratch.
A team at the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) and Northwestern University have demonstrated a new machine learning method that can design intrinsically disordered proteins with tailored properties. The work opens doors to new understanding of these mysterious biomolecules and possible new insights into origins of and treatments for diseases.
The work is published in Nature Computational Science and was co-led by SEAS graduate student Ryan Krueger and former NSF-Simons QuantBio Fellow Krishna Shrinivas, now an assistant professor at Northwestern, in collaboration with Michael Brenner, the Catalyst Professor of Applied Mathematics and Applied Physics at SEAS.
Shrinivas said he became interested in studying intrinsically disordered proteins because they are out of reach of current AI-based methods, such as Google DeepMind’s AlphaFold, for predicting and designing proteins with distinct conformations. Yet, such disordered proteins are important to many fundamental aspects of biology, and it is known that mutations to these proteins are linked to diseases like cancer and neurodegeneration. One example of a disordered protein is alpha-synuclein, long implicated in Parkinson’s and other diseases. To design IDPs for synthetic or therapeutics uses, Shrinivas said, “we needed to either come up with better AI models, or, we needed to come up with a way to actually take those physics models where you not only get good predictions, but you also get the physics for free.”
Automatic differentiation algorithms
The paper describes a computational method powered by algorithms that can perform “automatic differentiation,” or automatic computation of derivatives – instantaneous rates of change – in order to rationally select for protein sequences with desired behaviors or properties. The technique is a widely used tool for deep learning and training neural networks, but Brenner and his lab were among the first to recognize other potential use cases, such as optimizing physics-based molecular dynamics simulations.
With automatic differentiation, the researchers were able to train a computer to recognize how small changes in protein sequences – even single amino acid changes – affect the final desired properties of proteins. They likened their method to a very powerful search engine for amino acid sequences that fit the criteria needed to perform a function – say, one that creates loops or connectors, or can sense different things in the environment.
“We didn’t want to have to take a bunch of data and train a machine learning model to design proteins,” Krueger said. “We wanted to leverage existing, sufficiently accurate simulations to be able to design proteins at the level of those simulations.”
The method leverages a traditional framework for training neural networks called gradient-based optimization to identify new protein sequences with efficiency and precision. The result is that the proteins the researchers designed are “differentiable,” that is, they are not best-guesses predicted by AI, but rather based in molecular dynamics simulations, using real physics, that take into account how proteins actually, dynamically behave in nature.
The research received federal support from the National Science Foundation AI Institute of Dynamic Systems, the Office of Naval Research, the Harvard Materials Research Science and Engineering Center, and the NSF-Simons Center for Mathematical and Statistical Analysis of Biology at Harvard.
END
Order from disordered proteins
Physics-based machine learning algorithm applied to biomolecule design
2025-10-06
ELSE PRESS RELEASES FROM THIS DATE:
Rocket test proves bacteria survive space launch and re-entry unharmed
2025-10-06
A world-first study has proven microbes essential for human health can survive the extreme forces of space launch.
Space agencies are planning to send crews to Mars within decades but sustaining life on the red planet would be more difficult if important bacteria die during the flight.
Now an Australian-led study has found the spores of Bacilus subtilis, a bacterium essential for human health, can survive rapid acceleration, short-duration microgravity and rapid deceleration.
The spores of bacteria were launched high into ...
New wheat diversity discovery could provide an urgently-needed solution to global food security
2025-10-06
Wheat has a very large and complex genome. Researchers have found that different varieties can use their genes in different ways. By studying RNA—the molecules that carry out instructions from DNA—researchers can see which genes are active and when. By mapping this gene activity for the first time, researchers are able accelerate international wheat breeding programmes, developing new varieties of wheat which can adapt to the rapidly escalating climate emergency.
Wheat is the most widely cultivated crop in the world, with over 215 million hectares grown annually. To meet the demands of a growing global population, plant breeders face ...
Could reducing inflammation help combat fatigue in people with early-stage breast cancer?
2025-10-06
New research reveals that inflammatory responses may play a role in different types of fatigue experienced by many people with cancer. The findings are published by Wiley online in CANCER, a peer-reviewed journal of the American Cancer Society.
Cancer-related fatigue can be a distressing and persistent burden that causes patients to feel physical, emotional, and/or cognitive tiredness or exhaustion. Activation of inflammatory responses by the tumor itself and/or by cancer treatment is thought to be a key biological driver of this symptom, but inflammatory activity across the cancer continuum has not been thoroughly ...
Traumatic brain injuries in older adults linked to increased risk of dementia
2025-10-06
Traumatic brain injuries (TBI) in older adults are associated with new cases of dementia, use of home care services, and admission to long-term care, according to new research published in CMAJ (Canadian Medical Association Journal) https://www.cmaj.ca/lookup/doi/10.1503/cmaj.250361.
“One of the most common reasons for TBI in older adulthood is sustaining a fall, which is often preventable,” write Dr. Yu Qing Huang, ICES PhD student, and Dr. Jennifer Watt, associate professor and ICES scientist, both geriatricians at St. Michael’s ...
New intervention helps pediatricians promote early peanut introduction to prevent peanut allergy
2025-10-06
Results from a randomized clinical trial published in Pediatrics show that pediatricians who received targeted educational and clinical decision-support tools were significantly more likely to advise parents to introduce peanut-containing foods early – an approach recommended by national prevention guidelines but often underused in practice.
About 8% of U.S. children have food allergies. Peanut allergy, the most common pediatric food allergy, affects more than 2% of children nationwide.
“We found ...
New survey: Most Americans believe plasma donation saves lives, yet few have donated
2025-10-06
UNDER EMBARGO UNTIL OCT 6 at 12:01 a.m. ET
Media Contact:
media@pptaglobal.org
New Survey: Most Americans Believe Plasma Donation Saves Lives, Yet Few Have Donated
October 6 – Washington, D.C. – International Plasma Awareness Week (IPAW) takes place globally from October 6-10. The week encourages communities around the world to recognize the vital role of plasma in creating lifesaving medicines and honors those who donate. A new survey has found that 72% of respondents agree that plasma-derived medicines can save lives, yet most have not donated plasma.
The survey asked over 1,000 American adults questions about their understanding of plasma donation and lifesaving plasma-derived ...
New tools boost pediatricians’ adherence to peanut allergy guidelines 15-fold
2025-10-06
Doctors were nearly 15 times more likely to follow allergy-prevention guidelines with new tools
Tools included a training video and health record prompts for doctors and handouts for families
Randomized trial included 30 pediatric practices and 18,480 infant visits
CHICAGO --- A few easy-to-implement tools — a training video, electronic health record prompts and handouts for families — greatly increased how often pediatricians recommended early peanut introduction to infants, reports a new clinical study led by Northwestern University and Ann & Robert H. Lurie Children's Hospital of Chicago.
Since 2017, national guidelines have encouraged ...
Research unearths origins of Ancient Egypt’s Karnak Temple
2025-10-05
UNDER EMBARGO UNTIL 00:01AM UK TIME ON MONDAY 6 OCTOBER 2025
Research unearths origins of Ancient Egypt’s Karnak Temple
Most complete study of the temple complex and its landscape establishes earliest occupation and hints at link to creation myth
Researchers have carried out the most comprehensive geoarchaeological survey of Egypt’s Karnak Temple near Luxor – one of the ancient world’s largest temple complexes and a UNESCO World Heritage site welcoming millions of tourists every year.
The study, published in Antiquity today [6 October] reveals new evidence on the ...
Reevaluating nonoperative management for pediatric uncomplicated acute appendicitis
2025-10-05
About The Study: In contrast to earlier studies, this meta-analysis found significantly higher treatment failure and major complication rates within a year with nonoperative management among children and adolescents. The meta-analysis provides pediatricians and pediatric surgeons with up-to-date data to inform shared decision-making with families and encourage individualized, patient-centered treatment.
Corresponding Author: To contact the corresponding author, Isabella Faria, MD, email imdefrei@utmb.edu.
To access the embargoed ...
Metabolically active visceral fat linked to aggressive endometrial cancer, new study reveals
2025-10-04
(Barcelona, Spain, Sunday 5 October 2025) High metabolic activity of visceral fat may be associated with more aggressive endometrial cancer, new research presented today at the 38th Annual Congress of the European Association of Nuclear Medicine (EANM'25) has shown.1
While obesity is a recognised risk factor for endometrial cancer,2-3 the findings indicate that disease aggressiveness may be driven not only by the amount of visceral fat but also by its metabolic activity. Visceral fat is the fat that surrounds the internal organs, and to a larger degree than subcutaneous ...
LAST 30 PRESS RELEASES:
Pusan National University researchers reveal how forest soil properties influence arsenic mobility and toxicity in soil organisms
Korea University researchers find sweet taste cells resist nerve damage through c-Kit protein
HealthFORCE, AAPA, and West Health release “Aging Well with AI” – first in a two part series on AI and the healthcare workforce
The real reasons Endurance sank — study finds Shackleton knew of ship’s shortcomings
Marine heatwaves have hidden impacts on ocean food webs and carbon cycling
Order from disordered proteins
Rocket test proves bacteria survive space launch and re-entry unharmed
New wheat diversity discovery could provide an urgently-needed solution to global food security
Could reducing inflammation help combat fatigue in people with early-stage breast cancer?
Traumatic brain injuries in older adults linked to increased risk of dementia
New intervention helps pediatricians promote early peanut introduction to prevent peanut allergy
New survey: Most Americans believe plasma donation saves lives, yet few have donated
New tools boost pediatricians’ adherence to peanut allergy guidelines 15-fold
Research unearths origins of Ancient Egypt’s Karnak Temple
Reevaluating nonoperative management for pediatric uncomplicated acute appendicitis
Metabolically active visceral fat linked to aggressive endometrial cancer, new study reveals
Scientists glimpse how enzymes “dance” while they work, and why that’s important
California partnership aided COVID-19 response and health equity, report finds
University of Oklahoma secures $19.9 million for revolutionary radar technology
Study finds restoring order to dividing cancer cells may prevent metastasis
High-accuracy tumor detection with label-free microscopy and neural networks
Wayne State research reveals fetuses exposed to Zika virus have long-term immune challenges
Researchers deconstruct chikungunya outbreaks to improve prediction and vaccine development
Study finds one-year change on CT scans linked to future outcomes in fibrotic lung disease
Discovery of a novel intracellular trafficking pathway in plant cells
New tool helps forecast volcano slope collapses and tsunamis
Molecular coating cleans up noisy quantum light
From Parkinson's to rare diseases, discovered a key switch for cellular health
Tiny sugars in the brain disrupt emotional circuits, fueling depression
Mini-organs reveal how the cervix defends itself
[Press-News.org] Order from disordered proteinsPhysics-based machine learning algorithm applied to biomolecule design