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: | Anandita Nadkarni |
| Institution: | Carnegie Mellon Univeristy |
| Department: | Computational Biology |
| Country: | |
| Proposed Analysis: | I am part of the undergraduate team at CMU competing in project X (https://www.uoft.ai/projectx-2021). As such I do not have a BS (but I had to select something to be able to submit this form, please let me know if there's anything I can change to make that more accurate). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. To that end we would like access to PET and WGS data from ADNI, but we understand the first step is to try and seek authorization for an account which is what I have been tasked with trying to accomplish. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Additional Investigators | |
| Investigator's Name: | Alan Lai |
| Proposed Analysis: | Same as the proposed one that got approved, same deal with the degrees (currently seeking a bachelors, but not obtained). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Investigator's Name: | Hefei Tu |
| Proposed Analysis: | Same as the proposed one that got approved, same deal with the degrees (currently seeking a bachelors, but not obtained). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Investigator's Name: | Aarthi Ramsundar |
| Proposed Analysis: | Same as the proposed one that got approved, same deal with the degrees (currently seeking a bachelors, but not obtained). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Investigator's Name: | Yingjian Pan |
| Proposed Analysis: | Same as the proposed one that got approved, same deal with the degrees (currently seeking a bachelors, but not obtained). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Investigator's Name: | Shalin Shah |
| Proposed Analysis: | Same as the proposed one that got approved, same deal with the degrees (currently seeking a bachelors, but not obtained). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Investigator's Name: | Haohan Wang |
| Proposed Analysis: | This is the postdoc supervising us! Same as the proposed one that got approved, same deal with the degrees (currently seeking a bachelors, but not obtained). For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
| Investigator's Name: | Reid Simmons |
| Proposed Analysis: | This is the professor supervising us! For our competition project, our team would like to work on learning possible subtypes of Alzheimer's. We would like to analyze non-coding variants of Alzheimer's patients to investigate the role that somatic mutations may play in creating different pathways of AD. At our current stage of the competition (which is academic in nature), we hope to use subtypes that Vogel et. al discriminated from PET ADNI data (paper is Four distinct trajectories of tau deposition identified in Alzheimer’s disease) to reduce the above goal to a classification problem. If granted access to the data, we would plan to reproduce the labelings from Vogel et al. and then try to learn these subtypes from dimensionally reduced WGS data (specifically the non-coding SNVs, excluding all coding variants, as this has helped the accuracy of the classification of breast cancer into subtypes) using a random forest classifier. Depending on the success of that step we would want to proceed with either more sophisticated learning or more nuanced dimensionality reduction. We may still apply for the assembly data at a later date but for now just an authorized account should get us underway. Our team also has a post-doc and a professor supervising it (the post-doc has worked with ADNI data as part of a lab before, our team is separate and distinct from this effort and thus we are applying from scratch). |
