Tracing Alzheimer’s Genetic Footprints: A Pioneering Longitudinal Study Using Artificial Intelligence to Unravel Mutation- Driven Risks and Progression in Virtual Patients; Part 1 – The APOE genotypes

Abstract

Alzheimer's Disease (AD), the most common neurodegenerative disorder, presents a significant challenge for early detection and intervention due to its complex etiology involving genetic, environmental, and lifestyle factors. This study harnesses the innovative potential of artificial intelligence (AI) through the aiHumanoid platform to simulate AD progression. In this study, we focus on the impact of two APOE genotypes on disease development and progression. Our longitudinal virtual subject simulations, grounded in extensive medical literature and genetic information, explore the nuanced interplay between specific genetic variants, APOE ε3/4 and ε4/4, and their role in AD's heterogeneity. Despite the potential limitations associated with emerging technologies, including the translatability of AI simulations to real-world scenarios and the scope of genetic variants, this research provides key insights into early biomarkers and the progression patterns of AD. Future segments of this study (Part 2 and Part 3) will broaden the analysis to encompass a wider array of genetic factors and their interactions, enhancing the understanding of AD and paving the way for personalized intervention strategies. Ethical considerations surrounding the use of AI in medical research are acknowledged, emphasizing the need for responsible integration of technology in healthcare. Our findings underscore the transformative potential of AI in advancing AD research, offering a foundation for future studies aimed at refining diagnostic and therapeutic approaches through enhanced realism in simulations and a comprehensive exploration of genetic and environmental factors.

aiCRISPRL: An Artificial Intelligence Platform for Stem Cell and Organoid Simulation with Extensive Gene Editing Capabilities

“…The DeepNEU platform is an advanced literature-validated unsupervised machine learning platform [5][6][7][10][11][12][13][14][15][16][17][18][19][20][21]. Here we are presenting the updated database of DeepNEU (v8.4).…”

“…These four factors were given a value of +1 indicating that they were turned on for the first iteration and then allowed to evolve over successive iterations until a new system wide steady state is achieved. The detailed methodology for generating aiPSC simulations has been reported in detail previously [5][6][7][10][11][12][13][14][15][16][17][18][19][20][21].…”

“…Previous versions of the DeepNEU platform have successfully simulated a broad range of aiPSC derived cell types including neural stem cells, cardiac myocytes, skeletal muscle cells, lung cells, brain cells, ovarian cancer cells, lung adenocarcinoma cells, and natural killer cells (NK) [5][6][7][10][11][12][13][14][15][16][17][18][19][20][21]. Using a similar approach and a consensus gene mutational profile from multiple sources [22][23][24][25][26], we used the gene-editing capabilities of aiCRISPRL to create simulations (aiHeLa) that are most like original HeLa cells.…”

“…On review of our previous body of work [5][6][7][10][11][12][13][14][15][16][17][18][19][20][21], the aiCRISPRL functions of the DeepNEU platform successfully used gene editing to create single-gene mutations, multiple mutations and even edit the entire genome of the SARS-CoV-2 and NIPAH viruses. These edits created both loss of function (LOF) and gain of function (GOF) mutations.…”

Investigating Genetic and Familial Risks in Childhood ALL: A Longitudinal Virtual Study Using aiHumanoid Simulations of TEL-AML1 Gene Fusion

“…FCMs utilize fuzzy or grey scale logic to map causal relationships and their strength between various components of a system, such as genetic factors, environmental influences, and cellular pathways. This modeling technique which most resembles fully connected, recurrent artificial neural networks (RNN) has been successfully applied in various fields of medicine and is particularly suited for capturing the dynamic and multifactorial nature of diseases like ALL [7,8,9,10].…”