Machine learning reveals sex-specific Alzheimer's risk genes

NewsGuard 100/100 Score

A new study in the journal Nature Communications discusses the possibility of detecting genetic variants that could predict the risk of Alzheimer’s disease (AD) in males and females using a machine learning (ML) approach.

Study: Functional variants identify sex-specific genes and pathways in Alzheimer’s Disease. Image Credit: Kateryna Kon / Shutterstock.com Study: Functional variants identify sex-specific genes and pathways in Alzheimer’s Disease. Image Credit: Kateryna Kon / Shutterstock.com

Introduction

Less than one percent of AD cases arise due to dominantly inherited mutations. Sporadic late-onset AD (LOAD) is a major type of inherited AD, with 60-80% of the risk of LOAD attributable to genetic variation.

According to genome-wide association studies (GWAS), about one-third of the heritable risk of LOAD is from over 30 genetic loci. Most of these loci are related to the APOEε4 gene; however, many are located in non-coding parts of the genome, thus making it difficult to understand their contribution to the disease process.

Females suffer cognitive deterioration, brain shrinkage and lose more hippocampal neurons faster than males with the same APOE gene variants and age. However, males have a higher mortality rate from AD.

The higher risk of AD in females may be partially explained by the increased prevalence of depression, altered sleep, and cardiometabolic aberrations following the severe hormonal alterations of menopause. In addition, the presence of APOEε4 also increases the risk of AD in females.

Some sex-specific genes that have been shown to affect AD risk include ACE, BDNF, and REN.

It is critical to identify genetic contributors that underlie sex-differences in AD as it could lead to more accurate disease risk assessment and more tailored therapeutic approaches.”

What did the study show?

In the current study, researchers built an ML platform based on variants in the genetic code that have a possible functional impact on AD. The data was based on whole exome sequencing (WES) performed for the Alzheimer’s Disease Sequencing Project (ADSP).

WES data from 2,700 AD patients and 2,400 controls were used to identify functionally important non-synonymous coding variants of genes associated with AD. Most of these variants are of unknown significance; however, the evolutionary action (EA) score was used to estimate the potential impact of each variant on gene function.

Identification of AD genes

A total of 98 genes were identified, with APOE being the top-ranking gene. Thus, the EAML approach is capable of distinguishing genetic sequences from AD cases and those from controls, even with small datasets.

EAML genes were found to be expressed at higher levels in cellular pathways associated with AD. Of the 98, about half were also markedly dysregulated in AD as compared to controls, thus indicating their role in the neural damage associated with AD.

Further exploration identified 22 genes that are expressed at higher levels in immune response pathways in both sexes, some of which include cytokine signaling and microglial phagocytosis.

A total of 45 dysregulated genes were also enriched in immune response pathways. This supports earlier research indicating that neuroinflammation is a major contributor to AD through various genetic pathways.

Improved predictive ability

These genes were put into computational models intended to predict AD risk. The results showed an improvement in risk prediction with the use of these gene parameters as compared to that based on the age of onset and APOE variant.

ML predictors trained with gene features selected by EAML may have prognostic value for the risk of developing AD.”

EAML was also able to identify more than half of the candidate genes in smaller cohorts as compared to other models that only identified about 10% of genes. The top genes consistently maintained their ranking as cohort sizes decreased.

Most of these top genes have been identified as being dysregulated or contributing to AD pathogenesis. Thus, EAML appears to be a reliable predictive approach for AD risk, even at sample sizes that are unsuitable for other methods.

Neurodegeneration modifier genes

The findings were tested on the fruit fly Drosophila in two lab variants expressing abnormal tau or secreted Aβ42 protein. Seventy-three of the 98 genes were tested by introducing loss-of-function homologs.

Of these, 36 were found to modulate tau-related neurodegenerative pathways. More specifically, the reduced function of 24 genes led to slower neurodegeneration in the presence of tau, whereas neurodegeneration was accelerated with the remaining twelve loss-of-function variants.  

In the secreted Aβ42 model, 17 loss-of-function gene homologs were associated with enhanced neurodegeneration, whereas 12 were associated with improvement. Thus, 27 genes identified in this study reduced the adverse impact of neuronal impairment when knocked down. These findings thus provided in vivo validation of this platform.

Sex-specific genes

Sex-specific genes were identified when female and male sequences were processed separately. The ML approach showed that stress-response genes were emphasized in males, whereas cell cycle genes were higher in females. In fact, 21 of the overlapping 50 genes identified by a separate male-female analysis were identified by the combined analysis.

In females, five genes related to both AD and neuronal impairment caused by tau/Aβ42 were dysregulated in women with AD. This could reflect a higher AD risk in females than males, following impairment of cell cycle pathways.

What are the implications?

These proof-of-concept findings support a general approach to identify genetic mechanisms linked to complex diseases by machine learning over case-control sequence data using phylogenetic evolutionary information.”

The identification of 27 genes whose attenuation modulated neuronal deficits in Drosophila could be of therapeutic importance.

Several drugs that act on 11 of the genes identified in this study are currently available. In fact, one of the two male-specific genes is currently being evaluated in preclinical and clinical trials.

Of the four female-specific genes that have candidate drugs targeting their activity, some have shown beneficial effects in mouse models. Future sex-based clinical trials may help define which therapeutics are more effective in either sex.

In addition, the current study findings could help improve the accuracy of risk prediction based on genomic profiles, especially in males. Furthermore, the sex-specific mechanisms that contribute to AD could be better understood using EAML on larger datasets in the future.

A general approach for ML on functionally impactful variants can uncover sex-specific candidates towards diagnostic biomarkers and therapeutic targets.”

Journal reference:
  • Bourquard, T., Kwanghyuk, L., Al-Ramahi, I., et al. (2023). Functional variants identify sex-specific genes and pathways in Alzheimer’s Disease. Nature Communications. doi:10.1038/s41467-023-38374-z.
Dr. Liji Thomas

Written by

Dr. Liji Thomas

Dr. Liji Thomas is an OB-GYN, who graduated from the Government Medical College, University of Calicut, Kerala, in 2001. Liji practiced as a full-time consultant in obstetrics/gynecology in a private hospital for a few years following her graduation. She has counseled hundreds of patients facing issues from pregnancy-related problems and infertility, and has been in charge of over 2,000 deliveries, striving always to achieve a normal delivery rather than operative.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Thomas, Liji. (2023, May 16). Machine learning reveals sex-specific Alzheimer's risk genes. News-Medical. Retrieved on April 13, 2024 from https://www.news-medical.net/news/20230516/Machine-learning-reveals-sex-specific-Alzheimers-risk-genes.aspx.

  • MLA

    Thomas, Liji. "Machine learning reveals sex-specific Alzheimer's risk genes". News-Medical. 13 April 2024. <https://www.news-medical.net/news/20230516/Machine-learning-reveals-sex-specific-Alzheimers-risk-genes.aspx>.

  • Chicago

    Thomas, Liji. "Machine learning reveals sex-specific Alzheimer's risk genes". News-Medical. https://www.news-medical.net/news/20230516/Machine-learning-reveals-sex-specific-Alzheimers-risk-genes.aspx. (accessed April 13, 2024).

  • Harvard

    Thomas, Liji. 2023. Machine learning reveals sex-specific Alzheimer's risk genes. News-Medical, viewed 13 April 2024, https://www.news-medical.net/news/20230516/Machine-learning-reveals-sex-specific-Alzheimers-risk-genes.aspx.

Comments

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
Post a new comment
Post

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
MIT researchers develop new method to quickly screen cancer-associated genetic mutations