DeepoMe, a leading Generative AI for Longevity company, has announced the launch of their new foundation model for medicine platform, DamoPa. DamoPa is an abbreviation for Foundation Model for Medicine based on Pre-training.
DeepoMe, a leading Generative AI for Longevity company, has announced the launch of their new foundation model for medicine platform, DamoPa. DamoPa is an abbreviation for Foundation Model for Medicine based on Pre-training.
DamoPa is a collection of pre-trained model libraries and accompanying computing engines. Its main functions include generating biomarkers, disease prediction models, intervention targets, candidate drugs, and providing a computable framework for traditional Chinese medicine. An exciting feature of this system is that it can be used to characterize basic concepts and patterns in traditional Chinese medicine, making them measurable and quantifiable. The characterization of traditional Chinese medicine is essentially no different from that of cellular and molecular features, only at different levels of complexity.
The release of DamoPa demonstrates DeepoMe’s deep thinking about the future of medical models. DeepoMe’s goal is to explore a foundational data platform that combines the characteristics of traditional Chinese and Western medicine in an engineering way. Just as the chatGPT phenomenon indicates the value of “violent aesthetic” systems thinking, the development of medicine in the era of large-scale artificial intelligence models will also be driven by the “data flywheel.”
Currently, the knowledge systems used to characterize disease causes and mechanisms of drug action mainly consist of signal pathways manually edited by human experts. DamoPa aims to use a data-driven approach to calculate models related to human aging, disease etiology and mechanism, and drug action in large-scale datasets to form a pre-trained model library. In applications, based on local data (small-scale datasets, or even biological measurements of individual subjects), pre-trained models are applied to generate copies based on local data, and then test the correlation between the model and phenotype on local data. Or generate intervention plans based on endogenous metabolites, nutrients, and drugs.
Recently, a paper published in Genome Biology entitled “Ageing as a software design flaw” introduced a novel perspective. The traditional view suggests that aging is caused by accumulated damage to the body’s hardware, such as cell molecular damage caused by oxidative stress. However, this paper challenges this view and argues that aging is mainly caused by “software design flaws” in our bodies.
DamoPa uses epigenetic information, specifically DNA methylation, as the “data bus” for the entire system. For scientists working on DeepoMe, DNA methylation information is like a memo written during the operation of human software. DNA methylation information can reflect an individual’s historical environmental exposure and can also predict their future health characteristics and disease risks.
DeepoMe is planning to release a program that will showcase its data computing capabilities and serve as a collaborative platform for high-level demonstration applications in precision nutrition, traditional Chinese medicine, and anti-aging, especially in the field of traditional Chinese medicine.
DeepoMe’s scientists are at the forefront of developing innovative AI solutions that will revolutionize personalized medicine. To learn more about their perspective, technology, and products, please contact info@DeepoMe.com.