Biostate AI raises USD $12 million to expand RNAseq precision AI

Techday NZ

21-05-2025

Biostate AI has secured USD $12 million in a Series A funding round to expand its artificial intelligence technology for predicting human disease progression and drug response based on RNA sequencing data.
The investment round was led by Accel, with participation from Gaingels, Mana Ventures, InfoEdge Ventures, and previous supporters Matter Venture Partners, Vision Plus Capital, and Catapult Ventures. Biostate AI's earlier seed funding rounds included contributions from industry figures such as Dario Amodei, CEO of Anthropic, Mike Schnall-Levin, CTO of 10x Genomics, and Emily Leproust, CEO of Twist Bioscience.
The company intends to use the funds to advance affordable precision medicine solutions, particularly through RNA sequencing (RNAseq) services targeting molecular research in the United States. Biostate AI's immediate priorities are to increase the accessibility of RNAseq and to enhance the development of predictive clinical models, with an underlying mission to achieve personalised medical therapies.
Biostate AI was established by David Zhang and Ashwin Gopinath, both former academics and serial entrepreneurs. The founders recognised the potential for the entire RNA transcriptome to be used as a real-time biomarker for human health, addressing barriers of cost and analytics that have historically limited a comprehensive approach.
The company operates globally with sites in the US, India, and China, collaborating with hospitals, academic research teams, and life sciences partners. The ultimate aim is to construct what it describes as the world's largest RNA sequencing dataset to train general-purpose AI models capable of addressing a wide array of human diseases.
According to Biostate AI, existing RNA sequencing technology is commonly underused due to high costs, technical challenges with aggregating datasets across research sites, and inconsistencies introduced by varied laboratory methods—known as batch effects. These hurdles reduce opportunities for labs to scale projects, while also complicating fine-grained data analysis critical in clinical research.
To address these issues, Biostate AI has developed and patented technologies such as BIRT, which applies a multiplexing method for processing multiple tissue samples simultaneously, lowering costs and broadening tissue use from both fresh and archival sources. Another technology, PERD, is designed to distinguish signal from background noise in RNAseq datasets without reducing quality.
The economic efficiencies gained from these innovations allow Biostate AI to conduct internal experiments at reduced costs and to amass large volumes of de-identified RNAseq profiles. This extensive dataset is likened by the company to the way large language model AIs, such as those built by OpenAI, are trained on wide-ranging internet content.
The company also focuses on standardising workflows and data handling so that AI models can learn biological patterns more effectively, regardless of differences in sampling times or sites. This consistency is intended to limit confounding variables and enable reliable fine-tuning of foundational AI models for specific research applications.
An integrated software pipeline supports this approach, encompassing molecular sample processing, data aggregation, and clinical insight generation. The suite includes generative AI tools, such as Quantaquill, which translates analysis results into publication-ready manuscripts.
"Just as ChatGPT transformed language understanding by learning from trillions of words, we're learning the molecular language of human disease from billions of RNA expressions from millions of samples," said Ashwin Gopinath, co-founder and Chief Technology Officer of Biostate AI, and a former MIT assistant professor. "We're doing for molecular medicine what large language models did for text—scaling the raw data so the algorithms can finally shine."
Gopinath's drive for the company's work is closely linked to his wife's experience with leukaemia, and the co-founders' backgrounds in engineering and DNA research pointed them towards RNA as a less explored but valuable area for health insights. Their overall strategy is influenced by a self-sustaining business model inspired by subscription platforms.
Biostate AI's generative AI models seek to bridge the gap between molecular data and clinical intervention by identifying expression signatures in thousands of genes linked to various diseases and treatment responses. According to the company, this allows detection of subtle molecular changes that pre-empt the emergence of symptoms, potentially enabling much earlier medical response.
"Rather than solve the diagnostics and therapeutics as separate, siloed problems for each disease, we believe that the modern and future AI can be general purpose to understand and help cure every disease," said David Zhang, co-founder and Chief Executive Officer of Biostate AI, and former Associate Professor of Bioengineering at Rice University. "Every diagnostic I've built was about moving the answer closer to the patient. Biostate takes the biggest leap yet by making the whole transcriptome affordable."
The company has achieved initial proof-of-concept results in predicting disease recurrence for leukaemia patients and is planning further collaborations in oncology, autoimmune disorders, and cardiovascular disease. To date, Biostate AI reports it has processed over 10,000 RNAseq samples for more than 150 collaborating institutions and has arrangements to process hundreds of thousands of samples annually.
Biostate AI's client base includes over 100 pilot projects spanning various disease areas, such as leukaemia with Cornell University and multiple sclerosis with the Accelerated Cure Project. Total capital raised by the company now exceeds USD $20 million as it seeks to scale its AI model development and data collection efforts.