A new approach to diagnosing early-stage Alzheimer’s and Parkinson’s diseases has emerged from the AI biology firm Prima Mente. The company has developed an innovative model, known as Pleiades, that utilizes artificial intelligence to analyze changes in the epigenome, potentially transforming how neurodegenerative conditions are detected.
The challenge of diagnosing neurodegenerative diseases lies in their subtle onset, often beginning decades prior to the appearance of noticeable symptoms. Traditional methods, such as brain biopsies, are invasive and expensive. In contrast, Prima Mente’s approach analyzes cell-free DNA (cfDNA) obtained from patient blood samples, offering a less intrusive and cost-effective alternative.
Hannah Madan, PhD, co-founder of Prima Mente, articulated the significance of this new diagnostic pathway in an interview with GEN Edge. “Imaging and cognitive scores don’t necessarily tell you how to intervene or change the path of disease,” she explained. “This is what epigenetics can start to unravel.” The Pleiades model has already shown remarkable promise, achieving an impressive 89% success rate in detecting early-stage Alzheimer’s and Parkinson’s in real-world clinical cohorts. When combined with leading neurodegenerative protein biomarkers, such as pTau-217, the model’s accuracy increased to 97%.
Collaboration and Technology Advancements
Prima Mente, which was founded in 2022 and has its headquarters in London, also operates in San Francisco and Dubai. The company has joined the NVIDIA Inception program and partnered with Nebius, an NVIDIA Cloud Partner, to utilize the DGX Cloud Lepton AI platform for the training and development of Pleiades. Madan noted the importance of this collaboration, stating, “The hypothesis we had from the beginning was that bigger models would help understand complex biology over time, which requires both compute and the ‘know-how’ for building these models.”
Currently, Pleiades is undergoing testing in a live clinical study called SANDBOX, which spans 15 sites across the United Kingdom. The model is designed to assist in classifying and stratifying patients suspected of having neurodegenerative conditions, thereby improving clinical outcomes and managing waiting lists within the National Health Service.
The sophistication of the Pleiades model is underscored by its foundation on 7 billion parameters trained with 1.9 trillion tokens of epigenetic data. This data encompasses human methylation and genomic sequences, sourced from various cell types and sequenced in a manner that preserves cytosine methylation states. By using tissue-specific information, Pleiades can accurately identify the cfDNA’s cell type of origin, which includes brain components such as neurons and microglia, as well as immune components like T-cells and B-cells, all of which are potential predictors of Alzheimer’s disease.
Future Directions in Neurodegenerative Research
In addition to its diagnostic capabilities, Prima Mente is also developing a virtual cell model named Parthenon aimed at identifying new therapeutic targets related to microglia. Notably, Pleiades has revealed that cfDNA fragmentation patterns can serve as a new class of biomarkers for Alzheimer’s detection. These patterns are reflective of various biological processes, including epigenetic states and cell death mechanisms. The model has demonstrated an 83% accuracy rate in reconstructing cfDNA fragments and a 91% match to original methylation patterns. Furthermore, it enhances signals from specific blood cell types by up to 20-fold, facilitating the detection of rare disease signatures from limited sample sizes.
Looking ahead, Prima Mente plans to expand its research beyond epigenomics to include areas such as proteomics and transcriptomics. The company is actively conducting longitudinal studies that incorporate both blood samples and brain tissue. Madan has emphasized that the platform’s adaptability will allow it to be applied to a wider range of diseases by 2026.
As Prima Mente continues to innovate in the field of neurodegenerative diagnostics, its groundbreaking work has the potential to significantly impact patient care and treatment strategies, paving the way for a new era in precision medicine.
