US-based biotech company Triplebar has revealed plans to restructure its business to focus on the development of AI-driven models, to drive advancements in genomic research.

The firm’s CEO Maria Cho said that, as part of a strategic company initiative to leverage its capabilities to create proprietary genome-wide data linked to cellular function, Triplebar restructure to increase its focus on developing generative AI genomic language models.

Cho told The Cell Base: “The Triplebar platform today evolves animal cells to have specific properties that are required for scaling, including things like growing in suspension or low cost media, having fast doubling times, etc. The eventual goal would be to leverage a model to offer predictions for these kinds of phenotypes to make faster progress and to even identify other desirable traits through emergent properties of the model."

Cho explained how Triplebar is evolving cell lines with properties that can scale production efficiently. The company aims to bring cell-ag products to price parity or even below the cost of traditional agricultural methods, making them commercially viable. According to CEO Maria Cho, achieving this would be a crucial breakthrough for the industry, allowing cell-ag to compete directly with conventional food systems on cost and scale.

A major part of this strategy involves leveraging a generalised learning model (GLM) to streamline the optimisation of cell lines across various types. By using AI-driven technologies like a GLM, Cho believes that Triplebar can accelerate development and achieve significant advancements faster, unlocking widespread growth potential within the cell-ag space. This approach addresses a key challenge in the sector and could be pivotal in driving the industry forward.

“The time for AI as a tool in biology is now, and we are uniquely positioned to make a significant impact in this sector with our partners,” Cho said. “We are reorganising the business to position ourselves for this next phase, which involves leveraging our datasets to train genomic language models.”

Through the data sets Triplebar generates daily, in combination with the restructuring, Triplebar expects to meet its objective of developing the new AI genomic model by early 2026.

Cho likened Triplebar’s tech to natural language models like GPT4, or protein models like those from Alphafold. “Lesser known are genomic language models, which will be our focus,” she added.

Triplebar is already creating proteins identical to those found in nature at lower cost and less environmental damage.

#Triplebar #US