Survival Genie 2: a next-generation web server for targeted and single-cell-based survival analyses

Genome Med. 2026 May 6. doi: 10.1186/s13073-026-01651-9. Online ahead of print.

ABSTRACT

BACKGROUND: With the growing number of publicly available cancer genomics datasets, numerous web-based survival analysis tools have been developed to link gene expression data with patient outcomes, aiming to discover novel biomarkers and therapeutic targets. However, existing tools primarily support single-gene or gene-set analyses catered to targets derived from traditional sequencing approaches, limiting their applicability in the era of single-cell genomics.

RESULTS: Here, we introduce Survival Genie 2, the first web-based platform uniquely designed to facilitate survival analyses integrating user-generated single-cell data. Through this novel approach, our tool enables researchers to identify outcome-associated genes from single-cell cluster markers and weighted gene co-expression network modules. Furthermore, Survival Genie 2 offers specialized capabilities for parsing user inputs into biologically relevant categories, including long noncoding RNAs (lncRNAs), transcription factors, and ligand-receptor signaling genes, enhancing the robustness and interpretability of survival associations. Importantly, Survival Genie 2 also hosts the most comprehensive database to date, comprising 132 datasets from 53 cancer genome projects across 38 distinct malignancies, including both adult and pediatric cancers. The tool further provides versatile partitioning methods, generating analysis results through univariate Cox proportional hazards modeling, which are presented in publication-ready Kaplan-Meier and integrated forest plots. Additionally, Survival Genie 2 uniquely bridges the gap between publicly available bulk sequencing datasets and cell-type-specific targets derived from the user’s single-cell sequencing data by correlating user input gene expression with predicted immune cell enrichment, estimated through deconvolution analysis on the selected cancer dataset.

CONCLUSIONS: By integrating these innovative features, with specialized survival analyses on single-cell cluster markers, co-expression modules, regulatory networks, and ligand-receptor pairs, Survival Genie 2 represents a novel tool for expanding the translational impact of cancer research, enabling the precise identification of biomarkers and therapeutic targets. Survival Genie 2 is available at https://bhasinlab.bmi.emory.edu/SurvivalGenie2/home.

PMID:42092983 | DOI:10.1186/s13073-026-01651-9