Shares of Bruker Corporation and 10x Genomics moved lower on Monday following comments from Microsoft’s CEO about a new artificial intelligence model called GigaTIME. The model, developed in collaboration with Providence Health and the University of Washington, is designed to predict multiplex immunofluorescence results using standard H&E pathology slides.
The decline in these spatialomics stocks was largely driven by investor concerns. According to Wolfe Research, the market reacted to the possibility that AI-driven solutions could eventually reduce the need for spatial proteomics in clinical settings.
Despite this reaction, Wolfe Research suggested that the sell-off may be exaggerated. The firm highlighted that spatialomics technologies from Bruker and 10x Genomics are not yet widely adopted in clinical practice. In addition, the research behind GigaTIME was published in the journal Cell several months ago, indicating that the development is not entirely new.
GigaTIME uses a multimodal AI approach to convert traditional H&E pathology data into spatial proteomics insights. The study focused on 21 protein markers, including PDL-1, which plays a key role in immunotherapy. The model was trained on a dataset of 40 million cells and tested on more than 14,000 patients, combining both pathology slides and multiplex immunofluorescence data.
However, Wolfe Research pointed out that evaluating only 21 proteins may limit the model’s effectiveness compared to traditional methods, which typically analyze a much broader range of markers. For this reason, the firm considers the technology to still be in its early stages of development.
The tumor immune microenvironment remains a critical factor in cancer progression and response to treatment. While multiplex immunofluorescence is currently used to study this environment, it is often considered costly and time-consuming, which is one of the key challenges AI solutions like GigaTIME aim to address.
