- Rare variants with large effects provide functional insights into the pathology of migraine subtypes, with and without aura Nature.com
- Large Study of 80,000 Migraine Sufferers Uncovers Commonalities That May Lead to Potential Treatment Good News Network
- Migraines linked to rare genetic variants that could boost treatments New Scientist
- A large international study of migraine reveals new biological pathways for treatment Medical Xpress
- New migraine treatments could help ease misery for millions of Britons Express
- View Full Coverage on Google News
Tag Archives: functional
Groundbreaking achievement as bionic hand merges with user’s nervous and skeletal systems, remaining functional after years of daily use – EurekAlert
- Groundbreaking achievement as bionic hand merges with user’s nervous and skeletal systems, remaining functional after years of daily use EurekAlert
- It’s not ‘Star Wars’-level tech yet, but doctors get a step closer to a bionic hand with special surgery and AI CNN
- ‘Bionic woman’ is first to have robotic limb merged with bone — and controlled with her mind New York Post
- ‘Groundbreaking’ bionic arm that fuses with user’s skeleton and nerves could advance amputee care Euronews
- Bionic hand merges with user’s nervous and skeletal systems, remaining functional after years of daily use Medical Xpress
- View Full Coverage on Google News
Study: Donanemab Significantly Reduces Cognitive, Functional Decline Associated With Alzheimer Disease – Pharmacy Times
- Study: Donanemab Significantly Reduces Cognitive, Functional Decline Associated With Alzheimer Disease Pharmacy Times
- Alzheimer’s patient: New ‘miracle’ drug has ‘given me my life back’ New York Post
- New Alzheimer’s treatment: Donanemab drug seen as turning point in fight against disease FRANCE 24 English
- Eli Lilly says FDA approval of Alzheimer’s drug donanemab could come later this year Fox Business
- More developments being made in Alzheimer’s research WBAL-TV 11 Baltimore
- View Full Coverage on Google News
Precision functional MRI mapping reveals distinct connectivity patterns for depression associated with traumatic brain injury – Science
- Precision functional MRI mapping reveals distinct connectivity patterns for depression associated with traumatic brain injury Science
- Redefining Depression: TBI Affective Syndrome Discovered Neuroscience News
- Depression after a brain injury is a distinct condition, study finds. That could change how it’s treated. AOL
- Why depression after traumatic brain injury is distinct — and less likely to respond to standard treatment STAT
- Abnormal Brain Folding a Biomarker for Major Depressive Disorder Neuroscience News
- View Full Coverage on Google News
Cognitive functional therapy with or without movement sensor biofeedback versus usual care for chronic, disabling low back pain (RESTORE): a randomised, controlled, three-arm, parallel group, phase 3, clinical trial – The Lancet
- Cognitive functional therapy with or without movement sensor biofeedback versus usual care for chronic, disabling low back pain (RESTORE): a randomised, controlled, three-arm, parallel group, phase 3, clinical trial The Lancet
- Innovative therapy brings hope to chronic lower back pain sufferers New Atlas
- Study offers fresh hope for people living with chronic back pain Medical Xpress
- New drug-free treatment offers long-term hope for Australians struggling with back pain | 7NEWS 7NEWS Australia
- Curtin University researchers say back pain treatment trial gives hope to millions ABC News
- View Full Coverage on Google News
Pressure Chamber Therapy Is Effective in the Functional Improvement of Autism
Summary: Pressure chamber therapy significantly improved social skills and reduced inflammation in animal models of autism.
Source: Tel Aviv University
A new Tel Aviv University study succeeded in significantly improving social skills and the condition of the autistic brain through pressure chamber therapy.
The study was conducted on animal models of autism. In it, the researchers identified changes in the brain, including a reduction in neuroinflammation, which is known to be associated with autism.
Moreover, a significant improvement was found in the social functioning of the animal models treated in the pressure chamber. The study’s success has many implications regarding the applicability and understanding of treating autism using pressure chamber therapy.
The breakthrough was made under the leadership of doctoral student Inbar Fischer, from the laboratory of Dr. Boaz Barak of Tel Aviv University’s Sagol School of Neuroscience and School of Psychological Sciences. The research was published in the International Journal of Molecular Sciences.
Fischer and Barak explain that hyperbaric medicine is a form of therapy in which patients are treated in special chambers where the atmospheric pressure is higher than the pressure we experience at sea level, and in addition are delivered 100% oxygen to breathe.
Hyperbaric medicine is considered safe, and is already being used to treat a long list of medical conditions, including here in Israel.
In recent years, scientific evidence has been accumulating that unique protocols of hyperbaric treatments improve the supply of blood and oxygen to the brain, thereby improving brain function.
Dr. Barak says that “the medical causes of autism are numerous and varied, and ultimately create the diverse autistic spectrum with which we are familiar. About 20% of autistic cases today are explained by genetic causes, that is, those involving genetic defects, but not necessarily ones that are inherited from the parents.”
“Despite the variety of sources of autism, the entire spectrum of behavioral problems associated with it are still included under the single broad heading of ‘autism,’ and the treatments and medications offered do not necessarily correspond directly to the reason why the autism developed.”
In the preliminary phase of the study, a girl carrying the mutation in the SHANK3 gene, which is known to lead to autism, was treated by Prof. Shai Efrati, director of the Sagol Center for Hyperbaric Medicine at the Shamir “Assaf Harofeh” Medical Center, faculty member at the Sagol School of Neuroscience, and a partner in the study. Upon completing a series of treatments in the pressure chamber, it was evident that the girl’s social abilities and brain function had improved considerably.
In the next stage, and in order to comprehend the success of the treatment more deeply, the team of researchers at Dr. Barak’s laboratory sought to understand what being in a pressurized chamber does to the brain.
To this end, the researchers used adult animal models carrying the same genetic mutation in the SHANK3 gene as that carried by the girl who had been treated. The experiment comprised a protocol of 40 one-hour treatments in a pressure chamber, which lasted several weeks.
Dr. Barak says that they “discovered that treatment in the oxygen-enriched pressure chamber reduces inflammation in the brain and leads to an increase in the expression of substances responsible for improving blood and oxygen supply to the brain, and therefore brain function. In addition, we saw a decrease in the number of microglial cells, immune system cells that indicate inflammation, which is associated with autism.
“Beyond the neurological findings we discovered, what interested us more than anything was to see whether these improvements in the brain also led to an improvement in social behavior, which is known to be impaired in autistic individuals,” adds Dr. Barak.
“To our surprise, the findings showed a significant improvement in the social behavior of the animal models of autism that underwent treatment in the pressure chamber compared to those in the control group, who were exposed to air at normal pressure, and without oxygen enrichment.”
“The animal models that underwent treatment displayed increased social interest, preferring to spend more time in the company of new animals to which they were exposed in comparison to the animal models from the control group.”
Inbar Fischer concludes that “the mutation in the animal models is identical to the mutation that exists in humans. Therefore, our research is likely to have clinical implications for improving the pathological condition of autism resulting from this genetic mutation, and likely also of autism stemming from other causes.”
See also
“Because the pressure chamber treatment is non-intrusive and has been found to be safe, our findings are encouraging and demonstrate that this treatment may improve these behavioral and neurological aspects in humans as well, in addition to offering a scientific explanation of how they occur in the brain.”
About this autism research news
Author: Press Office
Source: Tel Aviv University
Contact: Press Office – Tel Aviv University
Image: The image is in the public domain
Original Research: Open access.
“Hyperbaric Oxygen Therapy Alleviates Social Behavior Dysfunction and Neuroinflammation in a Mouse Model for Autism Spectrum Disorders” by Inbar Fischer et al. International Journal of Molecular Sciences
Abstract
Hyperbaric Oxygen Therapy Alleviates Social Behavior Dysfunction and Neuroinflammation in a Mouse Model for Autism Spectrum Disorders
Autism spectrum disorder (ASD) is a multifactorial neurodevelopmental disorder (NDD) characterized by impaired social communication and repetitive behavior, among other symptoms. ASD is highly heritable, with SHANK3 being one of the high-risk genes for ASD.
In recent years, knowledge has been growing regarding the neuroplasticity effect induced by hyperbaric oxygen therapy (HBOT) and its potential use for ASD. Here, we characterized the effect of HBOT on a mouse model for ASD with the human genetic condition of InsG3680 mutation in the Shank3 gene. As compared to placebo, HBOT improved social behavior and reduced neuroinflammation in the cortex of the InsG3680(+/+) mice.
Specifically, HBOT induced upregulation of Insulin-like growth factor 1 (Igf1) expression levels and reduced the number of Iba1-positive cells in the mouse model for ASD compared to placebo control.
Together, our research suggests that HBOT has the potential to improve the clinical outcome of ASD by ameliorating some of the core pathophysiological processes responsible for the development of the disorder.
After their father and grandfather died of Alzheimer’s, these brothers began exploring how functional and psychedelic mushrooms can impact brain health
-
Chris and Joe Claussen watched their grandfather, then their father, die of Alzheimer’s disease.
-
They did genetic testing that found they’re at increased risk for the disease too.
-
They founded First Person, a company focused on the medicinal power of mushrooms.
When Chris and Joe Claussen were kids, they saw their father struggle when his own dad was diagnosed with Alzheimer’s disease. Years later, at just 65 years old, their dad was diagnosed with the condition that affects memory, brain function, and behavior.
“We saw how it progressed from little things to complete disaster,” Joe said.
The brothers had seen how the disease ends and they scrambled for answers. They were intrigued by the possible medicinal applications of mushrooms. They implemented dietary changes for their dad and saw small improvements in his abilities. Unfortunately, it was too late to reverse the course of the disease.
“The simple truth is, currently, if you wait too long to start taking care of your brain, there is not a lot you can do,” Chris said. “You have to start 20 to 30 years before you start to see symptoms.”
The brothers are at increased risk of Alzheimer’s
The brothers decided to do genetic testing to better understand their risk for Alzheimer’s. It revealed that they both carry the APOE4 gene, which doubles or triples the risk of developing Alzheimer’s. For the brothers, knowing that provided valuable information.
“It shouldn’t be scary, it should be a guide,” Chris said.
They started following a healthy-brain protocol that included following a ketogenic diet and supplementing with medicinal mushrooms like Lion’s Mane. Both said they noticed physical and cognitive improvements in themselves. That convinced them that they were onto something, and they started First Person.
“We wanted to get the message out there: The time to take care of your brain starts now,” Chris said.
Exploring the impact of mushrooms and microdosing
First Person manufactures supplements based on functional mushrooms to support brain health. Functional mushrooms like Lion’s Mane don’t contain psychoactive compounds that are banned in the US.
But the Claussens believe that those compounds are just as important in unlocking brain health. First Person is conducting research on psychedelic mushrooms in Jamaica, where the mushrooms are legal. Both the Claussens believe that some legalization in the US is inevitable, and they want their research to help support that.
“We’re studying these compounds and looking at standardizing dosing,” Joe said.
These are compounds that have been used for millennia, particularly in Asian cultures, Joe said. US drug policy has stigmatized them, but now Western researchers are starting to understand their potential. At the same time, there’s a strong grassroots movement, with everyone from moms to veterans extolling the benefits of microdosing — the practice of using very small amounts of psychedelics.
The brothers want to avoid disaster for their own families
The Claussens believe that despite their genetic risk, their futures are not sealed. Through dietary and lifestyle changes, they believe that Alzheimer’s can be prevented. Research shows they could be onto something: Up to 40% of dementia cases could be prevented through lifestyle adjustments like reducing alcohol intake and maintaining healthy blood pressure.
“This is just making a few small tweaks that are not overly difficult to do,” Chris said.
That can help your entire family avoid heartbreak in the future. “It affects everybody if you can’t take care of yourself,” Chris said.
Read the original article on Insider
Persistent loss of smell due to COVID-19 may better predict long-term cognitive, functional impairment
New insights into factors that may predict, increase or protect against the impact of COVID-19 and the pandemic on memory and thinking skills were revealed by multiple studies reported today at the Alzheimer’s Association International Conference® (AAIC®) 2022 in San Diego and virtually.
Among the key findings reported at AAIC 2022:
- A group from Argentina found that persistent loss of the sense of smell may be a better predictor of long-term cognitive and functional impairment than severity of the initial COVID-19 disease.
- Hospitalization in the intensive care unit was associated with double the risk of dementia in older adults, according to a study by Rush Alzheimer’s Disease Center in Chicago.
- During the pandemic, female gender, not working and lower socioeconomic status were associated with more cognitive symptoms in a large study population drawn from nine Latin American countries.
- In that same Latin American population, experiencing a positive life change during the pandemic (such as more quality time with friends and family or spending more time in nature) reduced the negative impact of the pandemic on memory and thinking skills.
“COVID-19 has sickened and killed millions of people around the world, and for some, the emerging research suggests there are long-term impacts on memory and thinking as well,” said Heather M. Snyder, Ph.D., vice president of medical and scientific relations at the Alzheimer’s Association. “As this virus will likely be with us for a long time, identifying the risk and protective factors for cognitive symptoms can assist with the treatment and prevention of ‘long COVID’ moving forward.”
Persistent loss of smell better predicts cognitive impairment than severity of COVID-19
Researchers in Argentina working with the Alzheimer’s Association Consortium on Chronic Neuropsychiatric Sequelae of SARS-CoV-2 Infection followed 766 adults age 55-95 exposed to COVID-19 for one year, and conducted a series of regular physical, cognitive and neuropsychiatric tests. Of the study group, 88.4% were infected and 11.6% were controls.
Clinical assessment showed functional memory impairment in two-thirds of the infected participants, which was severe in half of them. Another group of cognitive tests identified three groups with decreased performance:
- 11.7% showed memory-only impairment.
- 8.3% had impairment in attention and executive function.
- 11.6% displayed multidomain (including memory, learning, attention and executive function) impairment.
Statistical analysis revealed that persistent loss of smell was a significant predictor of cognitive impairment, but severity of the initial COVID-19 disease was not.
The more insight we have into what causes or at least predicts who will experience the significant long-term cognitive impact of COVID-19 infection, the better we can track it and begin to develop methods to prevent it.”
Gabriela Gonzalez-Aleman, LCP, Ph.D., Professor, Pontificia Universidad Catolica Argentina, Buenos Aires
A stay in the intensive care unit may signal higher dementia risk
Researchers from the Rush Alzheimer’s Disease Center (RADC), part of Chicago’s Rush University System for Health, used data from five diverse studies of older adults without known dementia (n=3,822) to observe intensive care unit (ICU) hospitalizations. ICU hospitalizations were previously linked to cognitive impairment in older patients, but few studies have examined whether they increase risk for dementia.
They reviewed Medicare claims records from 1991 to 2018 (pre-pandemic), and checked annually for development of Alzheimer’s and all type dementia using a standardized cognitive assessment. During an average 7.8 years follow up, 1,991 (52%) participants experienced at least one ICU hospitalization; 1,031 (27%) had an ICU stay before study enrollment; and 961 (25%) had an ICU stay during the study period.
The researchers found that, in analyses adjusted for age, sex, education and race, experiencing ICU hospitalization was associated with 63% higher risk of Alzheimer’s dementia and 71% higher risk of all type dementia. In models further adjusted for other health factors such as vascular risk factors and disease, other chronic medical conditions, and functional disabilities, the association was even stronger: ICU hospitalization was associated with 110% greater risk of Alzheimer’s and 120% greater risk of all type dementia.
“We found that ICU hospitalization was associated with double the risk of dementia in community-based older adults,” said Bryan D. James, Ph.D., epidemiologist at RADC. “These findings could be significant given the high rate of ICU hospitalization in older persons, and especially due to the tremendous upsurge in ICU hospitalizations during the COVID-19 pandemic. Understanding the link between ICU hospitalization and the development of dementia is of utmost importance now more than ever.”
“More research is necessary to replicate these findings and elucidate the factors that may increase dementia risk. For example, is it the critical illness that sends someone to the hospital or potentially modifiable procedures during the hospitalization that drives dementia risk?” James added.
One positive life change during the pandemic may buffer against cognitive symptoms
Investigators from countries across Central and South America and the United States examined whether sociodemographic factors and changes in life associated with the pandemic were related to experiencing cognitive symptoms, including problems with memory, attention and other thinking skills, during the early phases of the pandemic.
In the study reported at AAIC, 2,382 Spanish-speaking adults age 55-95 (average 65.3 years, 62.3% female) from nine countries in Latin America completed an online or telephone survey, had electronic cognitive testing, and filled out an inventory assessing the positive and negative impacts of the pandemic between May and December 2020. Of the total study population, 145 (6.09%) experienced COVID-19 symptoms.
Participants were from: Uruguay (1,423, 59.7%), Mexico (311, 13.1%), Peru (153, 6.4%), Chile (152, 6.4%), Dominican Republic (117, 4.9%), Argentina (106, 4.5%), Colombia (50, 2.1%), Ecuador (39, 1.6%), Puerto Rico (19, 0.8%) and Other (12, 0.5%)
Key findings:
- Female gender, not currently working and lower socioeconomic status were all independently associated with more cognitive symptoms during the early part of the pandemic.
- Negative life changes during the pandemic, such as economic difficulties and limited social activities, were significantly associated with more cognitive symptoms. However, this association was weaker among study participants who reported at least one positive life change during the pandemic, including spending more time with friends and family or more time outside in nature.
“Identifying risk and protective factors for cognitive symptoms during the pandemic is an important step towards the development of prevention efforts,” said María Marquine, Ph.D., associate professor in the Departments of Medicine and Psychiatry, and director of disparities research in the Division of Geriatrics, Gerontology and Palliative Care at the University of California, San Diego. “The experience of positive life changes during the pandemic might buffer the detrimental impact of negative life changes on cognitive symptoms.”
“This study is an example of how investigators from diverse countries in Latin America and the United States, many of whom had never worked together before and had limited resources, came together under difficult circumstances but with a shared goal to advance scientific understanding about Alzheimer’s, and the important contributions that such multicultural partnerships can yield,” Marquine added.
Heightened dream recall ability linked to increased creativity and functional brain connectivity
People who can frequently recall their dreams tend to be more creative and exhibit increased functional connectivity in a key brain network, according to new research published in the journal Nature and Science of Sleep. The findings provide new insights into the neurophysiological correlates of dreaming.
“I think that dreaming is one of the last frontiers of human cognition — a terra incognita of the mind if you will,” said study author Raphael Vallat, a postdoctoral researcher at the Center for Human Sleep Science at the University of California, Berkeley. “Although we all spend a significant amount of our lives dreaming, there are still so many basic research questions related to dreams that are unanswered, which obviously makes it such a fascinating topic to study!
“In this and previous studies, we address one of these fundamental research questions: why do some people recall their dreams every day while others almost never seem to recall a dream?”
For his new study, Vallat and his colleagues used brain imaging techniques to examine whether neurophysiological differences exist between individuals who frequently recall their dreams and those who do not.
The study included 55 healthy participants (ages 19–29) with normal sleep characteristics and body mass index. Twenty-eight participants were high dream recallers (able to recall about 6 dreams per week on average), while 27 participants were low dream recallers (recalling less than one dream per week on average). The two groups did not significantly differ in age, habitual sleep duration, or education.
Participants arrived at the sleep lab at Le Vinatier Hospital the night before their scanning session and completed self-reported assessments of personality, anxiety, and sleep quality. They also completed the Wechsler Memory Scale (used to measure immediate and delayed memory performance), the Guildford Uses Task (used to measure creative ability), and a digit span task (used to measure working memory’s number storage capacity). After staying at the lab overnight, the participants underwent three functional magnetic resonance imaging scans to measure resting-state brain activity.
The researchers found that high dream recallers and low dream recallers had similar personalities, levels of anxiety, sleep quality, and memory abilities. However, high dream recallers scored significantly higher on the Guildford Uses Task than low dream recallers, indicating that they had greater creative abilities.
Vallat and his colleagues also observed increased functional connectivity within the default mode network in high dream recallers compared to low dream recallers. The brain network “is known to be active during day-dreaming, mind-wandering (e.g. getting lost in your thoughts), and has been further suggested to promote creativity and dreaming,” Vallat explained. The increased connectivity was specifically found between the medial prefrontal cortex and the temporo-parietal junction, in line with clinical reports that have shown lesions to these brain regions result in a cessation of dream recall.
“In simpler words, high dream recallers have superior creative abilities, as well as a different brain functional organization, as demonstrated by this study and previous studies from our lab,” Vallat told PsyPost. “It remains an open question whether there is a causal relationship between dream recall, creative thinking, and brain ‘wiring’, and if so, what is the direction of that relationship (the chicken or egg problem). Does increased dreaming promote creative thinking and ultimately lead to changes in brain function? Or does an innate higher functional connectivity of the default mode network in these individuals promote their dream recall and creative abilities?”
An experimental methodology could help to untangle the causal relationships. “A next step of this study could be to take a group of non-dreamers, increase their dream recall abilities over time using some validated methods (the most known of which is to simply write down their dreams every morning as they wake up, the conscious effort of remembering their dreams eventually leading to a better recall of dreams), and assess their creativity and brain function before and after the manipulation,” Vallat explained.
But the study, like all research, includes some limitations. “Like most functional magnetic resonance imaging (fMRI) studies, we have used a fairly small sample size, which limits the generalizability of our findings (i.e. do these findings hold for a larger and more diverse population?),” Vallat said.
The study also only examined one type of creativity. In the Guildford Uses Task, participants are given two minutes to list as many alternative uses as possible for an everyday object. The total number of responses and the number of rare uses are used to measure a type of creative ability known as divergent thinking. “Creativity is an umbrella term that encompasses several concepts (e.g. convergent vs divergent thinking, problem solving, gist extraction, etc). In this study, we have measured a single subdomain of creativity,” Vallat noted.
“Understanding differences in dream recall between individuals is just one angle through which we are trying to decipher this fascinating and mysterious phenomenon that is dreaming,” Vallat said. “Studying dreams is a nightmare (sorry for the pun!) because it is not directly observable: we do not know exactly when dreaming happens during sleep, and we must therefore rely on waking up the sleeper to ask whether they were dreaming or not prior to awakening. Even then, this is imperfect because if they do not report any dreams, we cannot know for sure whether they were not dreaming or were in fact dreaming but immediately forgot their dream(s) upon waking.”
The study, “High Dream Recall Frequency is Associated with Increased Creativity and Default Mode Network Connectivity“, was authored by Raphael Vallat, Başak Türker, Alain Nicolas, and Perrine Ruby.
for(var key in aepc_pixel_args) args[key] = aepc_pixel_args[key];
return args; };
// Extend args if ( 'yes' === aepc_pixel.enable_advanced_events ) { aepc_pixel_args.userAgent = navigator.userAgent; aepc_pixel_args.language = navigator.language;
if ( document.referrer.indexOf( document.domain ) < 0 ) {
aepc_pixel_args.referrer = document.referrer;
}
}
!function(f,b,e,v,n,t,s){if(f.fbq)return;n=f.fbq=function(){n.callMethod?
n.callMethod.apply(n,arguments):n.queue.push(arguments)};if(!f._fbq)f._fbq=n;
n.push=n;n.loaded=!0;n.version='2.0';n.agent="dvpixelcaffeinewordpress";n.queue=[];t=b.createElement(e);t.async=!0;
t.src=v;s=b.getElementsByTagName(e)[0];s.parentNode.insertBefore(t,s)}(window,
document,'script','https://connect.facebook.net/en_US/fbevents.js');
fbq('init', aepc_pixel.pixel_id, aepc_pixel.user);
setTimeout( function() {
fbq('track', "PageView", aepc_pixel_args);
}, aepc_pixel.fire_delay * 1000 );
Read original article here
Study evaluates deep learning models that decode the functional properties of proteins
Deep learning–based language models, such as BERT, T5, XLNet and GPT, are promising for analyzing speech and texts. In recent years, however, they have also been applied in the fields of biomedicine and biotechnology to study genetic codes and proteins.
Bioinformaticians, genetics researchers and neuroscientists have been trying to infer the biological roles of genes and proteins for decades. To do this, however, they need to analyze extremely large and highly complex biological data.
Researchers at Hacettepe University, Middle East Technical University and Karadeniz Technical University, Turkey, have recently carried out a study evaluating the potential of deep learning–based language models for studying proteins and predicting their functional properties. Their paper, published in Nature Machine Intelligence, provides a valuable summary of the advantages and disadvantages of different state-of-the-art approaches.
“Molecular biology data can be modeled as a language (i.e., the language of genes/proteins), such that the sequence of a gene or protein can be thought of as a sentence with a specific meaning in natural language, and the semantics of this protein language is the specific biological, physical and chemical properties of these biomolecules,” Tunca Doğan, one of the researchers who carried out the study, told Phys.org. “Based on this idea, our work tries to build machine learning models that take language model-derived high dimensional numerical embeddings of proteins as input and predict their functional properties with high accuracy.”
In their paper, Doğan and his colleagues assessed the ability of different protein language modeling approaches to extract hidden patterns containing important clues about the functional properties of proteins. Their evaluations included all the most well-known natural language modeling architectures (i.e., BERT, T5, XLNet, ELMO, etc.), each of which can contain hundreds of millions or, in some cases, billions of parameters.
“Self-supervised pre-training of these models requires huge resources,” Doğan explained. “Thanks to valuable previous work on this topic, which aimed to pre-train protein language models using these architectures, we mostly focused on our secondary supervised training for predicting functional properties.”
In order to effectively assess the protein language models and compare their performances, the team first had to compile large and reliable testing datasets, each with a different difficulty level. Ultimately, they created four benchmark datasets that allowed them to investigate semantic similarities, ontology-based functional definitions, drug target protein families, and physical interactions between proteins. All of these are crucial biological mechanisms that are known to be closely linked to the occurrence and progression of genetically inherited diseases, such as different types of cancer.
“Perhaps our most notable finding was that these deep language models are able to successfully learn the functional properties of proteins using the amino acid sequences as the sole input, which is quite a difficult problem,” Doğan said. “These results are also consistent with the findings of recent protein structure prediction studies (e.g., Deepmind’s AlphaFold2 and Baker Lab’s RoseTTAFold), which uses the sequence as its input and predicts the 3D monomer structure with extremely high performance.”
In the future, the models evaluated by this team of researchers could help to enhance precision medicine interventions, for instance analyzing the molecular make-up of patients resulting from genomic variations to devise personalized treatments. While the results gathered by Doğan and his colleagues highlight the huge potential of deep learning–based protein modeling tools, existing methods will still need to be significantly improved before they can be integrated into real-life clinical decision-making systems.
“We are now working on a new system to better represent proteins,” Doğan added. “In addition to amino acid sequences, this system utilizes network-based data (i.e., known protein-protein interactions) and knowledge hidden in the unstructured biomedical texts (e.g., scientific articles) at the input level, together with integrative deep learning approaches. Our ultimate aim is to obtain a universal protein representation that can successfully be used in any biomedical or biotechnological modeling task.”
Researchers identify new medicines using interpretable deep learning predictions
More information:
Serbulent Unsal et al, Learning functional properties of proteins with language models, Nature Machine Intelligence (2022). DOI: 10.1038/s42256-022-00457-9
Serbulent Unsal et al, Learning functional properties of proteins with language models, Nature Machine Intelligence (2022). DOI: 10.1038/s42256-022-00457-9
© 2022 Science X Network
Citation:
Study evaluates deep learning models that decode the functional properties of proteins (2022, April 18)
retrieved 20 April 2022
from https://phys.org/news/2022-04-deep-decode-functional-properties-proteins.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.