Deciphering universal gene regulatory mechanisms in diverse organisms holds great potential for advancing our knowledge of fundamental life processes and facilitating clinical applications. However, the traditional research paradigm primarily focuses on individual model organisms and does not integrate various cell types across species. Recent breakthroughs in single-cell sequencing and deep learning techniques present an unprecedented opportunity to address this challenge. In this study, we built an extensive dataset of over 120 million human and mouse single-cell transcriptomes. After data preprocessing, we obtained 101,768,420 single-cell transcriptomes and developed a knowledge-informed cross-species foundation model, named GeneCompass. During pre-training, GeneCompass effectively integrated four types of prior biological knowledge to enhance our understanding of gene regulatory mechanisms in a self-supervised manner. By fine-tuning for multiple downstream tasks, GeneCompass outperformed state-of-the-art models in diverse applications for a single species and unlocked new realms of cross-species biological investigations. We also employed GeneCompass to search for key factors associated with cell fate transition and showed that the predicted candidate genes could successfully induce the differentiation of human embryonic stem cells into the gonadal fate. Overall, GeneCompass demonstrates the advantages of using artificial intelligence technology to decipher universal gene regulatory mechanisms and shows tremendous potential for accelerating the discovery of critical cell fate regulators and candidate drug targets.

Deciphering the universal gene regulatory mechanisms in diverse organisms holds great potential to advance our knowledge of fundamental life process and facilitate research on clinical applications. However, the traditional research paradigm primarily focuses on individual model organisms, resulting in limited collection and integration of complex features on various cell types across species. Recent breakthroughs in single-cell sequencing and advancements in deep learning techniques present an unprecedented opportunity to tackle this challenge. In this study, we developed GeneCompass, the first knowledge-informed, cross-species foundation model pre-trained on an extensive dataset of over 120 million single-cell transcriptomes from human and mouse. During pre-training, GeneCompass effectively integrates four types of biological prior knowledge to enhance the understanding of gene regulatory mechanisms in a self-supervised manner. Fine-tuning towards multiple downstream tasks, GeneCompass outperforms competing state-of-the-art models in multiple tasks on single species and unlocks new realms of cross-species biological investigation. Overall, GeneCompass marks a milestone in advancing knowledge of universal gene regulatory mechanisms and accelerating the discovery of key cell fate regulators and candidate targets for drug development.

Emerging single‐cell sequencing technology has generated large amounts of data, allowing analysis of cellular dynamics and gene regulation at the single‐cell resolution. Advances in artificial intelligence enhance life sciences research by delivering critical insights and optimizing data analysis processes. However, inconsistent data processing quality and standards remain to be a major challenge. Here scCompass is proposed, which provides a comprehensive resource designed to build large‐scale, multi‐species, and model‐friendly single‐cell data collection. By applying standardized data pre‐processing, scCompass integrates and curates transcriptomic data from nearly 105 million single cells across 13 species. Using this extensive dataset, it is able to identify stable expression genes (SEGs) and organ‐specific expression genes (OSGs) in humans and mice. Different scalable datasets are provided that can be easily adapted for AI model training and the pretrained checkpoints with state‐of‐the‐art single‐cell foundation models. In summary, scCompass is highly efficient and scalable database for AI‐ready, which combined with user‐friendly data sharing, visualization, and online analysis, greatly simplifies data access and exploitation for researchers in single‐cell biology ( http://www.bdbe.cn/kun ).

The awakening of dormant disseminated cancer cells appears to be responsible for the clinical relapses of patients whose primary tumors have been successfully cured months and even years earlier. In the present study, we demonstrate that dormant breast cancer cells lodged in the lungs reside in a highly mesenchymal, nonproliferative phenotypic state. The awakening of these cells is not triggered by a cancer cell-autonomous process. Instead, lung inflammation induced by the chemotherapeutic agent bleomycin effectively awakens dormant cancer cells, providing useful models for studying metastatic awakening. Mechanistically, the awakened cells shift from a highly mesenchymal to a quasi-mesenchymal phenotypic state in which they acquire tumorigenicity and proliferative ability. Once awakened, these cells can stably reside in this quasi-mesenchymal state and maintain their tumor-initiating ability, doing so without ongoing heterotypic signaling from the lung microenvironment. Epidermal growth factor receptor ligands released by the cells of the injured tissue microenvironment, including notably M2 type macrophages, promote dormant cancer cells to move toward this quasi-mesenchymal state, a transition that is critical for the awakening process. An understanding of the mechanisms of metastatic awakening may lead in the future to treatment strategies designed to prevent such awakening and resulting metastatic relapse.