Official websites use .gov

A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS

A lock ( 🔒 ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Need Help

Ultrascale Machine Learning to Empower Discovery in Alzheimer’s Disease Biobanks

Award Number: U01AG068057
ORGANIZATION: NATIONAL INSTITUTE ON AGING
OPDIV: NIH
AWARD CLASS: COOPERATIVE AGREEMENT
AWARD ACTIVITY TYPE: SCIENTIFIC/HEALTH RESEARCH (INCLUDES SURVEYS)
PERIOD OF PERFORMANCE START DATE: 09/15/2020
PERIOD OF PERFORMANCE END DATE: 01/31/2031

Group Awards By:

View Award Description

Ultrascale Machine Learning to Empower Discovery in Alzheimer’s Disease Biobanks - ABSTRACT Our U01 renewal application - Artificial Intelligence for Alzheimer's Disease (AI4AD) - expands on four productive themes from our prior funding period, supported by strong empirical results in our recent high impact publications. By developing cutting-edge Artificial Intelligence and Machine Learning (AIML) approaches, we address four key NIA mandates: (1) molecular subtyping for precision medicine, (2) improving Alzheimer’s Disease and Related Dementias (ADRD) clinical trial design, (3) adapting AI models to diverse cohorts for genetic target and treatment selection, and (4) genome-guided drug repositioning. Aim 1 develops a genomic extension of Large-Language Models to discover genomic markers associated with clinical AD onset and progression. We will adapt this breakthrough technology to process genomic sequences, to discover genomic targets associated with Alzheimer’s Disease (AD), AD biomarkers, and endophenotypes. We will compare the findings to those from our other AI-driven tools for gene discovery (SWAT-CNN, DeepBlock) and our whole genome tiling approach. In the Alzheimer’s Disease Sequencing Project (ADSP), over 60k people have 3-billion base-pair whole genome sequences–linked to imaging and deep phenotyping. Our AI-driven Genomic Discovery toolbox will reveal new AD-relevant genomic motifs that improve AD classification and predict disease progression more accurately than traditional AD GWAS-based models, such as polygenic risk. Aim 2 focuses on ADRD subtyping. Our AI tools will unravel the heterogeneity of dementia by deriving subtypes based on (1) clinical and cognitive data, (2) molecular neuropathology, and (3) neuroimaging. Subtyping across multiple dimensions will help more precisely define a patient’s diagnosis, and will eventually enhance treatment efficacy, re-stratify drug trial data to identify best responders, and allow gene discovery focused on subtypes. Our pilot work trained an AI method on ADSP and NACC datasets with post mortem neuropathology to infer scores for TDP-43, cerebral amyloid angiopathy, and Lewy body mixed-pathology subtypes. After validating this method on the large, diverse, incoming ADSP cohorts, we will run GWAS and whole-genome tiled association analyses to find genomic signatures linked to these subtypes. Aim 3 ensures our AI-based models are applicable across all ancestries, addressing the lack of validation for non-European groups. Our ancestry-informed models will leverage the new wave of ADSP data on local ancestry and social/environmental determinants of health. Aim 4 will enhance genome-guided treatment of ADRD by developing PreSiBo, an end-to-end drug discovery system matching treatments to specific subtypes. Focusing on molecular subtyping, we will validate subtype-specific targets and drugs using medication data from large cohorts, data from iPSC lines, and brain samples, based on the finding that ADRD drugs with genomic support are more likely to achieve FDA approval. Our renewal will advance precision medicine in ADRD through AI- driven approaches, creating robust. inclusive strategies for diverse populations.


Issue Date FY	Funding FY	Legal Entity Name	Legal Entity Address	Legal Entity City	Legal Entity State	Legal Entity Zip Code	Legal Entity COUNTY	Legal Entity COUNTRY	Assistance Listing	Award Code	Budget Year	Action Date	Action Type	Action Amount

Issue Date FY: 2026 ( Subtotal = $2,564,449 )
2026	2026	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90033	LOS ANGELES	USA	Aging Research	001	6	2/10/2026	COMPETING CONTINUATION	$2,564,449
2026	2026	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90089	LOS ANGELES	USA	Aging Research	002	6	2/27/2026	COMPETING CONTINUATION	$0
2026	2024	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90033	LOS ANGELES	USA	Aging Research	000	5	2/2/2026	NON-COMPETING CONTINUATION	$0
														Subtotal = $2,564,449

Issue Date FY: 2025 ( Subtotal = $0 )
2025	2024	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90033	LOS ANGELES	USA	Aging Research	000	5	3/19/2025	NON-COMPETING CONTINUATION	$0
2025	2024	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90033	LOS ANGELES	USA	Aging Research	001	5	8/20/2025	NON-COMPETING CONTINUATION	$0
														Subtotal = $0

Issue Date FY: 2024 ( Subtotal = $3,353,561 )
2024	2024	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90033	LOS ANGELES	USA	Aging Research	001	5	9/5/2024	NON-COMPETING CONTINUATION	$0
2024	2024	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90033	LOS ANGELES	USA	Aging Research	000	5	8/22/2024	NON-COMPETING CONTINUATION	$3,353,561
														Subtotal = $3,353,561

Issue Date FY: 2023 ( Subtotal = $3,747,255 )
2023	2023	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST	LOS ANGELES	CA	90089	LOS ANGELES	USA	Aging Research	003	4	8/25/2023	SUPPLEMENT FOR EXPANSION	$357,549
2023	2023	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST	LOS ANGELES	CA	90089	LOS ANGELES	USA	Aging Research	002	4	8/18/2023	NON-COMPETING CONTINUATION	$3,389,706
2023	2022	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST	LOS ANGELES	CA	90089	LOS ANGELES	USA	Aging Research	001	3	8/18/2023	NON-COMPETING CONTINUATION	$0
2023	2022	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	000	3	11/2/2022	NON-COMPETING CONTINUATION	$0
														Subtotal = $3,747,255

Issue Date FY: 2022 ( Subtotal = $3,641,568 )
2022	2022	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	002	3	9/2/2022	NON-COMPETING CONTINUATION	$3,641,568
2022	2021	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	000	2	11/9/2021	NON-COMPETING CONTINUATION	$0
2022	2021	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	001	2	7/18/2022	NON-COMPETING CONTINUATION	$0
														Subtotal = $3,641,568

Issue Date FY: 2021 ( Subtotal = $3,671,617 )
2021	2021	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	001	2	8/25/2021	NON-COMPETING CONTINUATION	$3,671,617
2021	2020	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	000	1	11/4/2020	NEW	$0
														Subtotal = $3,671,617

Issue Date FY: 2020 ( Subtotal = $3,767,729 )
2020	2020	UNIVERSITY OF SOUTHERN CALIFORNIA	3720 S FLOWER ST FL 3	LOS ANGELES	CA	90007	LOS ANGELES	USA	Aging Research	000	1	9/14/2020	NEW	$3,767,729
														Subtotal = $3,767,729

Grand Total All Awards = $20,746,179

Top

All Categories

About

Search

Reports

Data Submission

Award Information

Ultrascale Machine Learning to Empower Discovery in Alzheimer’s Disease Biobanks

Award Number: U01AG068057

ORGANIZATION: NATIONAL INSTITUTE ON AGING

OPDIV: NIH

AWARD CLASS: COOPERATIVE AGREEMENT

AWARD ACTIVITY TYPE: SCIENTIFIC/HEALTH RESEARCH (INCLUDES SURVEYS)

PERIOD OF PERFORMANCE START DATE: 09/15/2020

PERIOD OF PERFORMANCE END DATE: 01/31/2031

Federal Websites

Department of Health & Human Services

HHS Operating Divisions

HHS Staff Divisions

Download A Document Viewer