scPrognosis

"scPrognosis" is a computational method to enhance breast cancer prognosis using single-cell RNA-sequencing (scRNA-seq) data. Addressing the critical challenge of tumor heterogeneity, which significantly impacts prognosis accuracy, scPrognosis focuses on leveraging the insights provided by scRNA-seq data to uncover the complex cellular diversity within tumors. The method explicitly utilizes data related to the Epithelial-to-Mesenchymal Transition (EMT), a biological process known for its role in cancer progression and metastasis.

Key Features and Functionalities:

- EMT Pseudotime and Dynamic Gene Co-expression Network Inference: scPrognosis begins by inferring the EMT pseudotime of individual cells and constructing a dynamic gene co-expression network. This technique allows for a nuanced understanding of the EMT process at different stages.

- Integrative Model for Gene Selection: An integrative model is used to select genes important for EMT based on their expression variation and differentiation across EMT stages and their roles in the dynamic gene co-expression network. This selection process ensures that the identified genes are biologically relevant to breast cancer progression.

- Application to Breast Cancer Prognosis: The signature genes selected by scPrognosis are then used as features to build a prediction model using bulk RNA-seq data. This step bridges the gap between single-cell level insights and practical applications for prognosis using more readily available bulk RNA-seq datasets.

- Insights into EMT and Clinical Outcomes: Beyond its prognostic capabilities, scPrognosis offers valuable insights into the dynamic changes in gene expression associated with EMT and how these changes may correlate with clinical outcomes in breast cancer patients.