4. Alzheimer's disease is one of the most prevalent neurodegenerative diseases;
however, the genetic causes, such as specific mutations or an accumulation of
mutations, still evade the scientific community's consensus. The mitochondria are
essential in producing energy, and so dysfunction will lead to energy-related problems
and abnormal functioning of neuronal cells; thus, studying mutations found in the
mitochondria is essential. Data was collected from the Alzheimer's Disease
Neuroimaging Initiative and National Center of Biotechnology Information databases,
and machine learning algorithms analyzed the mutations found in both the patient
and control groups. In the H haplogroup, there was a separation between the two
groups; however, when further examined in the sub-haplogroups, only H4 showed
separation, while the others had the clustering of patient and control groups together.
The study aimed to create a method for predicting whether an unknown person
belongs in the control or patient group by examining their mitochondrial mutations.
Abstract
5. Introduction
• Alzheimer’s disease is one of the most prevalent neurodegenerative
diseases.
• Possible genetic answer:
Mitochondrial dysfunction
• Hypothesis:
Mitochondrial mutations may contribute to an individual’s
development of Alzheimer’s disease.
• Goal:
To identify mitochondrial markers
associated with Alzheimer’s disease
and use these mutations to determine
if an individual possesses this disorder.
https://medicine.umich.edu/dept/mneuronet/news/archive/201907/early-
alzheimer%E2%80%99s-disease-detection-may-benefit-new-stem-cell-therapy
6. Methodology/Experimental
• Data collected from the Alzheimer’s Disease Neuroimaging Initiative (ADNI)
and the National Center of Biotechnology Information (NCBI) database.
https://blog.23andme.com/ancestry-reports/haplogroups-explained/
Patient Example
Control Example
Python Scripts
7. Methodology/Experimental
• The mutations were accumulated from the control and patient groups, and
then analyzed through machine learning software analysis:
Principal Component Analysis (PCA),
Support Vector Machine learning (SVM), &
Linear Discriminant Analysis (LDA).
…
11. Conclusions
• Haplogroup H – there is some separation between the control and patient
group.
• Machine learning analysis works for some sub-haplogroups and does
not work for others.
• Mitochondrial mutations may be sub-haplogroup specific.
• Results due to differing sequencing and alignment methods.
• Future research:
• Tabulate mutations that occur more in the control or more in the
patient group.
• Study different haplogroups to see if a similar pattern is present.
12. Acknowledgments
• Special thanks to the members of the Hao Lab and especially Dr. Weilong
Hao for all the hours of assistance!
• Alzheimer’s Disease Neuroimaging Initiative for access to the database.