BrainStorm Cell Therapeutics news
Creating Waves: An AI Startup Accelerates Disease Research by Including a Lab. Neurological disorders, which vary from hundreds of rare diseases to Parkinson’s and Alzheimer’s, affect over a billion people, or 15% of the global population.
Patient-derived stem cells form organoids, miniature, three-dimensional brain bundles. BrainStorm Therapeutics, a San Diego-based startup, is employing AI-powered computational drug discovery in conjunction with lab studies to speed up the development of treatments for these illnesses. Lab in the loop is a hybrid, iterative approach that speeds up drug development by utilising clinical data and AI models to inform one another.
“The brain is the last frontier in modern biology,” stated Robert Fremeau, the founder and CEO of BrainStorm. Prior to this, Fremeau served as a neuroscience scientific director at Amgen and taught at Duke University and the University of California, San Francisco. It can now begin to decipher the intricate biology of disease networks by fusing the capabilities of generative AI with the organoid disease models.
The company wants to find treatments that may be used for a variety of illnesses and reduce the over 93% failure rate of drug candidates for brain ailments during clinical trials. If these objectives are met, the development of treatments for both common and uncommon illnesses will proceed more quickly and profitably.
The incapacity of conventional preclinical models using rats or 2D cells to forecast human efficacy is the primary cause of this startlingly high clinical trial failure rate. The ones creating a platform that more accurately represents the intricacy of human neurobiology and raises the possibility of therapeutic success by fusing AI-driven analysis with human-derived brain organoids.
According to Fremeau and Yin, BrainStorm’s platform might shorten development times, lower R&D expenses, and greatly raise the likelihood that patients would receive successful treatments.
The NVIDIA BioNeMo Framework, a collection of programming tools, libraries, and models for computational drug development, was used to create the AI models used by BrainStorm Therapeutics that operate on NVIDIA GPUs in the cloud. The business is a part of the global network of innovative startups known as NVIDIA Inception.
Clinical Trial
In order to find promising targets for possible medications and clinical biomarkers, BrainStorm Therapeutics employs AI algorithms to create gene maps of brain illnesses. They can assess the efficacy of possible treatments before beginning clinical trials by using organoids to directly screen hundreds of drug compounds on human brain cells every day.
Brain waves can be detected by a scan such as an electroencephalogram (EEG), which gauges the electrical activity of neurones in the brain. Because their organoids also produce spontaneous brain waves, humans are able to replicate the intricate activity found in the human brain in a much smaller system. For the purpose of researching brain disorders, one can handle it similarly to a clinical experiment.
In order to find new drugs for Parkinson’s disease, which is linked to the death of neurones that make dopamine, a neurotransmitter that aids in movement and thought, BrainStorm Therapeutics is now employing organoids produced from patients.
Although several genetic variations in Parkinson’s disease cause failure in various cellular pathways, they all lead to the degeneration of dopamine neurones. It can find disease-modifying therapies that could delay, stop, or even reverse the progression of Parkinson’s disease by mapping and analyzing the biological impacts of these variations using AI models.
The BrainStorm team refined foundation models accessible through the BioNeMo Framework, such as the Geneformer model for gene expression analysis, using single-cell sequencing data from brain organoids. Patients having mutations in the GBA1 gene, the most prevalent genetic risk factor for Parkinson’s disease, provided the organoids.
Additionally, BrainStorm and the NVIDIA BioNeMo team are working together to maximize open-source access to the Geneformer model.
Increasing the Speed of Drug Discovery Research
BrainStorm’s patented technology allows it to replicate the biology of the human brain and model the potential effects of various treatments on a patient’s brain.
Also can rapidly reduce the number of therapy possibilities because this can be done thousands of times, much more swiftly, and much more affordably than in a wet lab. After that, one can use organoids to test the subset of medications that the AI model predicts will work. Everyone won’t test these medications on people unless they pass those tests.
Donepezil, a medication used to treat Alzheimer’s disease, may also be useful in treating Rett syndrome, a rare genetic neurodevelopmental disorder, due to this technology. The BrainStorm team was able to move from organoid screening to submitting an application for a phase 2 clinical study of the medication in Rett sufferers in just nine months. The U.S. Food and Drug Administration just approved this application.
Additionally, BrainStorm intends to create multimodal AI models that incorporate information from EEG scans, cell imaging, cell sequencing, and other sources.
To create the correct medications, you need high-quality, multimodal input data. By using this data to train AI models, scientists will be able to better understand disease, identify more effective medication candidates, and eventually identify predictive biomarkers for individual patients that will allow precision medicine to be delivered.
In order to detect CDKL5 Deficiency condition, another uncommon genetic neurodevelopmental condition, the company is working with the CURE5 Foundation to do the most thorough repurposed medication screen to date.
Research on rare diseases is evolving from a high-risk specialty to a vibrant field. By combining BrainStorm’s AI-powered organoid technology with NVIDIA’s accelerated computing resources and the NVIDIA BioNeMo platform, innovation is happening at a much faster rate and at a much lower cost. What used to take a decade and billions of dollars can now be studied in a few months using much less money.
