Illumina’s BioInsight AI Lab Wins 2026 Virtual Cell Challenge

Illumina’s BioInsight AI Lab won the 2026 Virtual Cell Challenge hosted by the Arc Institute, organizers announced.

The competition tested how well computer models predict cellular reactions to genetic perturbations in unseen cell types. Approximately 1,500 competitors entered this year’s contest.

Illumina secured first place by pairing data from its Billion Cell Atlas with a new cellular response prediction algorithm named CellularAI, according to a press release from Illumina. The winning algorithm matched real experimental results generated by the Arc Institute across multiple metrics when evaluated against hidden data.

Billion Cell Atlas Powers Winning Model

The Billion Cell Atlas captures how individual cells respond to genetic changes via CRISPR across more than 200 disease-relevant cell lines. Kyle Farh, vice president of artificial intelligence at Illumina, leads the BioInsight AI Lab.

Researchers are training these systems on massive biological datasets to simulate cellular responses and disease trajectories. Dr. Christina Theodoris, a computational biologist at the Gladstone Institutes in San Francisco, trained her GeneFormer model on 30 million cells and is currently scaling her latest training run to 175 million cells.

Dr. Alex Marson, a physician-scientist at the University of California, San Francisco, and his colleagues studied 22 million cells by engineering 12,000 sets of immune cells, shutting down a different gene in each set to observe the activity of 18,000 other genes.

Investment Surges for Virtual Cell Startups

Financial backing for biological modeling has surged alongside research interest. BioHub announced commitments of nearly $2 billion for data and modeling initiatives last month. GenBio.AI, a virtual-cell start-up co-founded in 2024 by Dr. Xing, joins companies such as Xaira Therapeutics—which has attracted significant investment—in building proprietary virtual cells.

“A lot of companies are jumping into this space,” Dr. Xing told DNYUZ.

“When I have a disease, you can basically do personalized drug trials based on my own virtual cell,” Dr. Xing said.

Other researchers aim to manipulate cell health directly. “We want to know how to nudge cells from less-healthy states to more-healthy states, or from older states to younger states,” Rory Stark, a computational biologist at Altos Labs in Los Angeles, told DNYUZ.

Skepticism Over Timeline and Hype

Dr. Anders expressed similar reservations about commercial claims. “It’s a good business model to make a big promise and hope that the investors are less able to judge whatever it is you’re saying than you are. I wouldn’t invest in them,” Dr. Anders said.

David Baker, a molecular biologist at the University of Washington who shared the 2024 Nobel Prize for protein design, notes that current models still miss many internal cellular dynamics. “We’re not really taking into account all the different things that are going on inside the cell,” Dr. Baker told DNYUZ, adding that virtual models could help test engineered immune cells designed to destroy cancer.

Illumina, founded in April 1998 and headquartered in San Diego, California, provides integrated systems for genetic analysis and serves customers in more than 155 countries. The company acquired Solexa in 2007 for approximately $650 million, gaining the sequencing-by-synthesis technology that became the foundation of its dominant platform; by 2014, Illumina machines accounted for more than 90 percent of all DNA data produced worldwide.

Meanwhile, a review published in Expert Opinion on Drug Discovery on Nov. 17, 2026, by Xixi Li and colleagues at the University of Melbourne and Helmholtz Munich said that predicting individual disease trajectories and treatment responses remains challenging in drug discovery because molecular profiling provides static snapshots that miss cellular dynamics, spatial context and adaptive responses, according to PubMed.