There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Younghee Lee |
| Institution: | University of Utah |
| Department: | Biomedical Informatics |
| Country: | |
| Proposed Analysis: | Aim 1: Assess if gene expression varies longitudinally by Braak/clinical stage. ADNI is in a unique position in which RNA microarray data has been collected longitudinally. In addition, a subset of the participants converted from cognitively normal to mild cognitive impairment (MCI) or Alzheimer’s disease (AD) dementia. This presents an opportunity to assess if blood transcriptomic signatures, such as differentially expressed genes (DEGs), are specific to certain clinical stages or if they are persistent throughout. This data can be used to understand possible disease causative components found in the vascular system. Furthermore, DEGs that are present during the preclinical stages of the disease may be potentially identified. If the DEG is persistent across all clinical stages for a subset of individuals that converted from cognitively normal to cognitive impairment when compared against controls, then this could be a potential early risk assessor or biomarker. Aim 2: Determine AD-specific DEGs using blood microarray meta-analysis Blood biomarkers is an ongoing field of research for neurodegenerative diseases. As such, we’d like to identify DEGs that are specific to Alzheimer’s disease by comparing and contrasting those found in ADNI with those found in other microarray studies for other common neurodegenerative diseases, such as Parkinson’s disease, Huntington’s disease, or frontotemporal dementia. There is a wide array of such data on the National Center for Biotechnology Information’s Gene Expression Omnibus (NCBI GEO) platform, many of which are case-control studies. By utilizing the blood gene expression data found in ADNI with others found in NCBI GEO, we can identify genetic factors or transcriptomic signatures that are unique to Alzheimer’s disease or shared across neurodegenerative disease. The purpose of this work is to potentially improve the specificity of blood biomarker candidates by focusing attention on those that are AD-specific. Aim 3: Develop machine-learning model predicting risk of conversion By utilizing our information of DEGs which are persistent across clinical and preclinical stages, as well as those that are specific to AD, we aim to develop a machine learning model that can take the transcriptomic and genomic information from the blood during the preclinical phase and predict the individual’s risk of Alzheimer’s disease. An approach such as this can become increasingly complex given different combinations of risk factors, such as age, sex, and APOE genotype. As such, we would aim to use unsupervised clustering methods to identify the DEGs with the most variation particular to these demographics, beginning with filtered DEGs identified in Aims 1 and 2. Then, we would use the most discriminant genes for each demographic and use these to predict risk of conversion using the baseline RNA expression as input and clinical conversion outcomes as targets in a supervised model. Previous studies have shown potential in unsupervised feature extraction and supervised prediction in cancer diagnosis, so we seek to extend this methodology to Alzheimer’s disease. The main goal of this work is to determine the efficacy of blood gene expression data in predicting the risk of Alzheimer’s disease development during the preclinical stages of the disease, and if a patient is ultimately at severe risk, then further treatment or preventative measures may be sought. |
| Additional Investigators | |
| Investigator's Name: | Joshua Mincer |
| Proposed Analysis: | Aim 1: Assess if gene expression varies longitudinally by Braak/clinical stage. ADNI is in a unique position in which RNA microarray data has been collected longitudinally. In addition, a subset of the participants converted from cognitively normal to mild cognitive impairment (MCI) or Alzheimer’s disease (AD) dementia. This presents an opportunity to assess if blood transcriptomic signatures, such as differentially expressed genes (DEGs), are specific to certain clinical stages or if they are persistent throughout. This data can be used to understand possible disease causative components found in the vascular system. Furthermore, DEGs that are present during the preclinical stages of the disease may be potentially identified. If the DEG is persistent across all clinical stages for a subset of individuals that converted from cognitively normal to cognitive impairment when compared against controls, then this could be a potential early risk assessor or biomarker. Aim 2: Determine AD-specific DEGs using blood microarray meta-analysis Blood biomarkers is an ongoing field of research for neurodegenerative diseases. As such, we’d like to identify DEGs that are specific to Alzheimer’s disease by comparing and contrasting those found in ADNI with those found in other microarray studies for other common neurodegenerative diseases, such as Parkinson’s disease, Huntington’s disease, or frontotemporal dementia. There is a wide array of such data on the National Center for Biotechnology Information’s Gene Expression Omnibus (NCBI GEO) platform, many of which are case-control studies. By utilizing the blood gene expression data found in ADNI with others found in NCBI GEO, we can identify genetic factors or transcriptomic signatures that are unique to Alzheimer’s disease or shared across neurodegenerative disease. The purpose of this work is to potentially improve the specificity of blood biomarker candidates by focusing attention on those that are AD-specific. Aim 3: Develop machine-learning model predicting risk of conversion By utilizing our information of DEGs which are persistent across clinical and preclinical stages, as well as those that are specific to AD, we aim to develop a machine learning model that can take the transcriptomic and genomic information from the blood during the preclinical phase and predict the individual’s risk of Alzheimer’s disease. An approach such as this can become increasingly complex given different combinations of risk factors, such as age, sex, and APOE genotype. As such, we would aim to use unsupervised clustering methods to identify the DEGs with the most variation particular to these demographics, beginning with filtered DEGs identified in Aims 1 and 2. Then, we would use the most discriminant genes for each demographic and use these to predict risk of conversion using the baseline RNA expression as input and clinical conversion outcomes as targets in a supervised model. Previous studies have shown potential in unsupervised feature extraction and supervised prediction in cancer diagnosis, so we seek to extend this methodology to Alzheimer’s disease. The main goal of this work is to determine the efficacy of blood gene expression data in predicting the risk of Alzheimer’s disease development during the preclinical stages of the disease, and if a patient is ultimately at severe risk, then further treatment or preventative measures may be sought. |
| Investigator's Name: | Senggyun Han |
| Proposed Analysis: | Aim 1: Assess if gene expression varies longitudinally by Braak/clinical stage. ADNI is in a unique position in which RNA microarray data has been collected longitudinally. In addition, a subset of the participants converted from cognitively normal to mild cognitive impairment (MCI) or Alzheimer’s disease (AD) dementia. This presents an opportunity to assess if blood transcriptomic signatures, such as differentially expressed genes (DEGs), are specific to certain clinical stages or if they are persistent throughout. This data can be used to understand possible disease causative components found in the vascular system. Furthermore, DEGs that are present during the preclinical stages of the disease may be potentially identified. If the DEG is persistent across all clinical stages for a subset of individuals that converted from cognitively normal to cognitive impairment when compared against controls, then this could be a potential early risk assessor or biomarker. Aim 2: Determine AD-specific DEGs using blood microarray meta-analysis Blood biomarkers is an ongoing field of research for neurodegenerative diseases. As such, we’d like to identify DEGs that are specific to Alzheimer’s disease by comparing and contrasting those found in ADNI with those found in other microarray studies for other common neurodegenerative diseases, such as Parkinson’s disease, Huntington’s disease, or frontotemporal dementia. There is a wide array of such data on the National Center for Biotechnology Information’s Gene Expression Omnibus (NCBI GEO) platform, many of which are case-control studies. By utilizing the blood gene expression data found in ADNI with others found in NCBI GEO, we can identify genetic factors or transcriptomic signatures that are unique to Alzheimer’s disease or shared across neurodegenerative disease. The purpose of this work is to potentially improve the specificity of blood biomarker candidates by focusing attention on those that are AD-specific. Aim 3: Develop machine-learning model predicting risk of conversion By utilizing our information of DEGs which are persistent across clinical and preclinical stages, as well as those that are specific to AD, we aim to develop a machine learning model that can take the transcriptomic and genomic information from the blood during the preclinical phase and predict the individual’s risk of Alzheimer’s disease. An approach such as this can become increasingly complex given different combinations of risk factors, such as age, sex, and APOE genotype. As such, we would aim to use unsupervised clustering methods to identify the DEGs with the most variation particular to these demographics, beginning with filtered DEGs identified in Aims 1 and 2. Then, we would use the most discriminant genes for each demographic and use these to predict risk of conversion using the baseline RNA expression as input and clinical conversion outcomes as targets in a supervised model. Previous studies have shown potential in unsupervised feature extraction and supervised prediction in cancer diagnosis, so we seek to extend this methodology to Alzheimer’s disease. The main goal of this work is to determine the efficacy of blood gene expression data in predicting the risk of Alzheimer’s disease development during the preclinical stages of the disease, and if a patient is ultimately at severe risk, then further treatment or preventative measures may be sought. |
| Investigator's Name: | Habtamu Aycheh |
| Proposed Analysis: | Aim 1: Assess if gene expression varies longitudinally by Braak/clinical stage. ADNI is in a unique position in which RNA microarray data has been collected longitudinally. In addition, a subset of the participants converted from cognitively normal to mild cognitive impairment (MCI) or Alzheimer’s disease (AD) dementia. This presents an opportunity to assess if blood transcriptomic signatures, such as differentially expressed genes (DEGs), are specific to certain clinical stages or if they are persistent throughout. This data can be used to understand possible disease causative components found in the vascular system. Furthermore, DEGs that are present during the preclinical stages of the disease may be potentially identified. If the DEG is persistent across all clinical stages for a subset of individuals that converted from cognitively normal to cognitive impairment when compared against controls, then this could be a potential early risk assessor or biomarker. Aim 2: Determine AD-specific DEGs using blood microarray meta-analysis Blood biomarkers is an ongoing field of research for neurodegenerative diseases. As such, we’d like to identify DEGs that are specific to Alzheimer’s disease by comparing and contrasting those found in ADNI with those found in other microarray studies for other common neurodegenerative diseases, such as Parkinson’s disease, Huntington’s disease, or frontotemporal dementia. There is a wide array of such data on the National Center for Biotechnology Information’s Gene Expression Omnibus (NCBI GEO) platform, many of which are case-control studies. By utilizing the blood gene expression data found in ADNI with others found in NCBI GEO, we can identify genetic factors or transcriptomic signatures that are unique to Alzheimer’s disease or shared across neurodegenerative disease. The purpose of this work is to potentially improve the specificity of blood biomarker candidates by focusing attention on those that are AD-specific. Aim 3: Develop machine-learning model predicting risk of conversion By utilizing our information of DEGs which are persistent across clinical and preclinical stages, as well as those that are specific to AD, we aim to develop a machine learning model that can take the transcriptomic and genomic information from the blood during the preclinical phase and predict the individual’s risk of Alzheimer’s disease. An approach such as this can become increasingly complex given different combinations of risk factors, such as age, sex, and APOE genotype. As such, we would aim to use unsupervised clustering methods to identify the DEGs with the most variation particular to these demographics, beginning with filtered DEGs identified in Aims 1 and 2. Then, we would use the most discriminant genes for each demographic and use these to predict risk of conversion using the baseline RNA expression as input and clinical conversion outcomes as targets in a supervised model. Previous studies have shown potential in unsupervised feature extraction and supervised prediction in cancer diagnosis, so we seek to extend this methodology to Alzheimer’s disease. The main goal of this work is to determine the efficacy of blood gene expression data in predicting the risk of Alzheimer’s disease development during the preclinical stages of the disease, and if a patient is ultimately at severe risk, then further treatment or preventative measures may be sought. |
