The Cancer Dependency Map (DepMap) has undergone a significant expansion through the integration of 147 genome-scale CRISPR screens performed on next-generation cancer models. This initiative, published in the journal Nature on August 5, 2026, aims to address long-standing limitations in precision oncology by incorporating organoids and spheroids that better reflect the biological complexity of human tumors.
Traditional cancer research has relied heavily on monolayer cell lines, which often fail to capture the full spectrum of genomic diversity found in patient populations. By profiling 314 next-generation models—including 237 carcinoma organoids and 77 central nervous system tumor spheroids—the research team identified critical mutations that were previously underrepresented in standard datasets. These include NRAS mutations in colorectal cancer and ESR1 alterations in breast cancer, alongside specific genomic markers in prostate cancer such as SPOP and RB1.
The study utilized the Celligner tool to assess how accurately these models recapitulate transcriptional cell states compared to patient-derived xenografts and traditional cell lines. Results indicated that next-generation models achieved a lineage inference accuracy of 69%, significantly outperforming traditional cell lines which reached only 35%. The researchers attributed the lower accuracy of traditional lines to their tendency to cluster in an undifferentiated state associated with epithelial–mesenchymal transition.
Genome-scale CRISPR screens were conducted using the Humagne enAsCas12a single-guide RNA library to identify gene essentiality across diverse cancer types. The team observed that screen quality in these 3D models was comparable to established datasets, with clear separation between essential and non-essential gene distributions. These findings provide a high-fidelity framework for identifying new therapeutic targets that remain hidden in conventional 2D culture systems.
Data integration was facilitated by the Chronos algorithm, which accounts for copy number effects and library-specific batch variations. This approach allowed the researchers to harmonize data across different growth formats and culture conditions, ensuring that the resulting resource remains consistent for downstream analysis. The study highlights that under serum-free conditions, the growth format—whether organoids or matrix-coated plates—does not significantly alter gene dependency profiles.
The preservation of transcriptional programs in these models offers a new pathway for discovering biomarker-linked vulnerabilities that are typically silenced in traditional cultures. By maintaining key features of the tumor microenvironment, these organoids allow for a more nuanced understanding of how specific gene dependencies are shaped by the cellular context. This integrated dataset is now available through the DepMap portal, providing the scientific community with a more comprehensive tool for drug target discovery.
The shift toward 3D modeling represents a move to bridge the gap between preclinical findings and clinical outcomes. By expanding the coverage of tumor subtypes, the project addresses a critical bottleneck in the development of personalized treatments for patients who currently lack targeted options. Future research will likely focus on refining these models to further improve their predictive power for clinical trial success.
The integration of these models into the broader DepMap infrastructure sets a new standard for functional genomics. As researchers continue to populate the portal with diverse patient-derived samples, the ability to map vulnerabilities across a wider array of molecular subtypes will likely accelerate the identification of candidates for future therapeutic intervention. The findings underscore the necessity of moving beyond traditional cell lines to capture the true molecular heterogeneity of human cancer.
This work demonstrates that the inclusion of organoids and spheroids is essential for the next phase of cancer dependency mapping. By successfully integrating these diverse models, the researchers have created a more versatile resource that accounts for the nuances of tumor growth and transcriptional states. The scientific community can now leverage this expanded dataset to refine the selection of targets for personalized medicine, potentially increasing the success rate of future clinical trials.
