The Tencent Life Science Laboratory, working in tandem with the First Affiliated Hospital of Guangzhou Medical University and the Guangzhou Institute of Respiratory Health, has successfully developed the DeepGEM pathological large-scale model. This marks a significant milestone in the realm of lung cancer gene mutation prediction. Leveraging just standard pathological section images, this model can execute mutation prediction within a mere minute, boasting an impressive accuracy range of 78% to 99%. Notably, it bypasses the need for gene sequencing. Instead, it harnesses AI technology to discern gene mutations by scrutinizing pathological images, thus dramatically cutting down on expenses and time. At its heart, the model excels in extracting 'morphological signals' from ordinary pathological images. It employs a multi-instance learning framework capable of directly processing entire images and generating spatial distribution maps of mutations. With robust adaptability, the model's prediction accuracy is on par with traditional methods. Looking ahead, the collaborators intend to construct a 'Pathological-Genetic Multimodal Large Model Platform' together. This initiative aims to broaden the application of AI technology across a wider array of cancer types and anatomical sites.