Biostate AI to Develop Predictive Models for Multiple Sclerosis Using Transcriptomic Data

Biostate AI, a startup building AI models from RNA sequencing data, has partnered with the Accelerated Cure Project (ACP) to develop next-generation predictive tools for multiple sclerosis (MS). 

The collaboration will use transcriptomic data from ACP’s biorepository—one of the largest MS-focused collections globally—to train transformer-based AI models aimed at forecasting disease progression and treatment response. The initiative marks Biostate AI’s entry into neuroimmunology and expands its portfolio of disease-specific generative models.

Founded in 2023 and co-located in Palo Alto and Houston, Biostate AI develops scalable RNA sequencing and bioinformatic analysis platforms to support high-throughput omics data generation. Its patented Barcode-Integrated Reverse Transcription (BIRT) technology enables profiling of all RNA species, including non-coding RNAs, at a significantly reduced cost. The company has also launched OmicsWeb Copilot, a conversational AI interface for analyzing RNAseq data, fine-tuned on thousands of proprietary datasets. These technologies are intended to support large-scale data generation and model training across a wide range of disease areas and biological questions.

Source: Biostate AI

Through the ACP collaboration, Biostate AI will sequence thousands of longitudinal patient samples using its BIRT platform to generate high-resolution RNA expression data. These will be used to train AI models capable of detecting molecular patterns linked to MS onset, relapse, remission, and response to therapy. According to ACP’s leadership, the partnership is expected to convert years of curated data into clinically meaningful insights. The models will aim to provide clinicians with tools for patient stratification and precision treatment selection, addressing the current limitations of trial-and-error approaches in MS care.

Biostate AI has raised over $4 million in funding from Matter Venture Partners, Vision Plus Capital, Catapult VC, and the Caltech Seed Fund. Individual backers include Dario Amodei (Anthropic), Emily Leproust (Twist Bioscience), Mike Schnall-Levin (10X Genomics), and Joris Poort (Rescale). 

The company has filed nine patents and in-licensed additional intellectual property from Caltech to expand its capabilities in multiomic analysis. Alongside academic and clinical collaborations, Biostate’s long-term objective is to build generative AI systems capable of modeling human and animal health changes, including toxicity and efficacy responses, at population scale.

Topics: AI & Digital