Deep biomarkers of human aging: Application of deep neural networks to biomarker development

Putin, Evgeny; Aliper, Alexander; Korzinkin, Mikhail; Moskalev, Alexey; Kolosov, Alexey; Ostrovskiy, Alexander; Cantor, Charles R.; Vijg, Jan; Zhavoronkov, Alex

doi:10.18632/aging.100968

Cited by 232 publications

(206 citation statements)

References 31 publications

(206 reference statements)

Supporting

Mentioning

201

Contrasting

Unclassified

Order By: Relevance

“…(The epidemiology of longevity is reviewed comprehensively elsewhere 23 .) Further insights into the links between ageing and neurodegeneration are being generated from genetic studies that explore not only longevity and exceptional lifespan but also the genetics of disease-free ageing 24,25 and the integration of genetics with other omics approaches 26 .…”

Section: Causes Of Brain Ageing and Neurodegenerationmentioning

confidence: 99%

Ageing, neurodegeneration and brain rejuvenation

Wyss-Coray

2016

Nature

652

483

View full text Add to dashboard Buy / Rent full text

Although systemic diseases take the biggest toll on human health and well-being, increasingly, a failing brain is the arbiter of a death preceded by a gradual loss of the essence of being. Ageing, which is fundamental to neurodegeneration and dementia, affects every organ in the body and seems to be encoded partly in a blood-based signature. Indeed, factors in the circulation have been shown to modulate ageing and to rejuvenate numerous organs, including the brain. The discovery of such factors, the identification of their origins and a deeper understanding of their functions is ushering in a new era in ageing and dementia research.

show abstract

“…The difficulty in the comparison of each of these studies is that different study cohorts have been used and that different parameters have been determined to be included in analyses. Previously, albumin has been determined as an important predictor of biological age and residual lifespan (37, 38), which should encourage cohorts to measure metabolomics platforms that include albumin levels. Ultimately, transcriptomics, proteomics, and metabolomics measures should be harmonized for human cohorts and then added to the models for biological age and physiological dysregulation and investigate generalizability over multiple studies.…”

Section: Discussionmentioning

confidence: 99%

Classification for Longevity Potential: The Use of Novel Biomarkers

Beekman

Heemst

et al. 2016

Front. Public Health

View full text Add to dashboard Buy / Rent full text

BackgroundIn older people, chronological age may not be the best predictor of residual lifespan and mortality, because with age the heterogeneity in health is increasing. Biomarkers for biological age and residual lifespan are being developed to predict disease and mortality better at an individual level than chronological age. In the current paper, we aim to classify a group of older people into those with longevity potential or controls.MethodsIn the Leiden Longevity Study participated 1671 offspring of nonagenarian siblings, as the group with longevity potential, and 744 similarly aged controls. Using known risk factors for cardiovascular disease, previously reported markers for human longevity and other physiological measures as predictors, classification models for longevity potential were constructed with multiple logistic regression of the offspring-control status.ResultsThe Framingham Risk Score (FRS) is predictive for longevity potential [area under the receiver operating characteristic curve (AUC) = 64.7]. Physiological parameters involved in immune responses and glucose, lipid and energy metabolism further improve the prediction performance for longevity potential (AUCmale = 71.4, AUCfemale = 68.7).ConclusionUsing the FRS, the classification of older people in groups with longevity potential and controls is moderate, but can be improved to a reasonably good classification in combination with markers of immune response, glucose, lipid, and energy metabolism. We show that individual classification of older people for longevity potential may be feasible using biomarkers from a wide variety of different biological processes.

show abstract

“…Due to the complexity of coordinated signalling and communication between cells wherein most events do not function in isolation, a systems level analysis based on transcriptomic, proteomic, and connectomic mechanisms is needed 15 . An omic analysis can be described as follow: macromic where big-data analytics reveal neural tract adaptation in response to injury and regeneration and system wide motor-sensory adaptation; mesomic where the connectome of newly formed synapses are mapped as structure-function models, revealing new circuitry between the regenerating axon and grafted substrates in the lesion site; micromic where the molecular identity of regenerating neurons and permissive substrates are established through the utilization of transcriptomic and proteomic network analysis.…”

Section: Methods Of Discoverymentioning

confidence: 99%

An omic approach to spinal cord injury mechanisms

Niekerk¹

2016

J Neurol Neuromedicine

View full text Add to dashboard Buy / Rent full text

Spinal cord injury (SCI) research continues to make substantial progress in identifying both neuron-intrinsic and neuron-extrinsic mechanisms that limit central nervous system (CNS) plasticity and regeneration. The identification of these mechanisms has in turn led to several novel strategies for therapeutically enhancing recovery of the injured CNS. Despite this progress, clinical translation remains a challenge for several reasons, including: 1) problems in projecting beneficial outcomes from small animal models to primate systems, 2) a lack of robust improvement in functional outcomes in animal models, and 3) difficulty replicating published reports in the field. Collectively, while the field has seen great progress, reconstructing the exquisite circuitry of the injured human CNS will require yet greater progress in both understanding of basic mechanisms underlying axonal growth and guidance, and testing of optimized therapies in models more predictive of potential human benefit.

show abstract

“…For example, the heart rate and QT interval of 15 children with type 1 diabetes were monitored overnight and used to accurately predict low blood glucose with a deep neural network model [78]. Aging.ai, which uses an ensemble of deep neural networks on 41 standardized blood test measurements, has been trained to predict an individual's chronological age [79].…”

Section: Self-diagnosis With Deep Learningmentioning

confidence: 99%

Computational biology: deep learning

Jones

Alasoo

Fishman

et al. 2017

Emerging Topics in Life Sciences

View full text Add to dashboard Buy / Rent full text

Deep learning is the trendiest tool in a computational biologist's toolbox. This exciting class of methods, based on artificial neural networks, quickly became popular due to its competitive performance in prediction problems. In pioneering early work, applying simple network architectures to abundant data already provided gains over traditional counterparts in functional genomics, image analysis, and medical diagnostics. Now, ideas for constructing and training networks and even off-the-shelf models have been adapted from the rapidly developing machine learning subfield to improve performance in a range of computational biology tasks. Here, we review some of these advances in the last 2 years.

show abstract

“…The continued development of new computational and machine learning methodologies will be required to countenance all the variables at play. Zhavoronkov and colleagues used neural networks trained on 60,000 patient blood samples and information obtained from routine health checks (Putin et al, 2016). They found that the ensemble of neural networks that performed best in predicting age-related outcomes were those that relied on a large combination of biomarkers to make their prediction.…”

Section: Biomarkers For Detecting Drug Efficacy Against Ageingmentioning

confidence: 99%

Therapeutic Manipulation of Ageing: Repurposing Old Dogs and Discovering New Tricks

Mallikarjun

Swift

2016

EBioMedicine

View full text Add to dashboard Buy / Rent full text

Ageing is a leading risk factor for many debilitating diseases. While age-related diseases have been the subject of over a century of intense investigation, until recently, physiological ageing was considered unavoidable. Pharmacological and genetic studies have since shown that ageing is a malleable process and that its abrogation can prevent its associated diseases. This review summarises a sample of the most promising efforts to deliver the products of ageing research to the clinic. Current efforts include the use of clinically approved drugs that have since been repurposed, as well as the development of novel therapeutics, to target ageing. Furthermore, ongoing research has sought reliable biomarkers of ageing that will accelerate the development of such therapeutics. Development of these technologies will improve quality of late-life and help relieve the enormous stress placed on state healthcare systems by a rapidly ageing global population. Thus, for both medical and socioeconomic reasons, it is imperative that ageing is made to yield to intervention.

show abstract

Deep biomarkers of human aging: Application of deep neural networks to biomarker development

Cited by 232 publications

References 31 publications

Ageing, neurodegeneration and brain rejuvenation

Classification for Longevity Potential: The Use of Novel Biomarkers

An omic approach to spinal cord injury mechanisms

Computational biology: deep learning

Therapeutic Manipulation of Ageing: Repurposing Old Dogs and Discovering New Tricks

Contact Info

Product

Resources

About