Tumour initiation requires phenotypic plasticity to override differentiation constraints and adopt proliferative, invasive behaviours. This Master’s project investigates the earliest cell-specific responses to oncogenic activation by analysing single-cell RNA-sequencing data from zebrafish basal epidermal and myeloid cells at 24 and 48 hours post-induction of HRASG12V. To this end, the data-driven method Stator was employed to identify transcriptional states based on higher-order gene interactions, independent of proximity in expression space. The results were compared against a hierarchical clustering approach integrating selective inference. Stator identified distinct transcriptional signatures corresponding to basal and superficial keratinocyte populations, as well as myeloid cells, from which neutrophils were notably abundant across the samples. A subset of basal keratinocytes in the HRASG12V condition exhibited a transcriptional program enriched for tumour-propagating markers lamc2 and cd44a, alongside genes upregulated in proliferation and metabolism. This preneoplastic (PNC) state was detected in both biological replicates at 24 hpi and was followed by a proliferative but lamc2- cd44a- state at 48 hpi, suggesting loss of cancer stem-like cell population. Comparison with hierarchical clustering showed agreement in the identification of the PNC population. Neutrophils were consistently enriched in the HRASG12V condition, indicating an associated immune response. These results demonstrate that Stator is able to resolve cell identity for the early detection of oncogenic reprogramming in a vertebrate model, providing one framework for investigating early tumorigenesis at single-cell resolution.
Tumour initiation requires phenotypic plasticity to override differentiation constraints and adopt proliferative, invasive behaviours. This Master’s project investigates the earliest cell-specific responses to oncogenic activation by analysing single-cell RNA-sequencing data from zebrafish basal epidermal and myeloid cells at 24 and 48 hours post-induction of HRASG12V. To this end, the data-driven method Stator was employed to identify transcriptional states based on higher-order gene interactions, independent of proximity in expression space. The results were compared against a hierarchical clustering approach integrating selective inference. Stator identified distinct transcriptional signatures corresponding to basal and superficial keratinocyte populations, as well as myeloid cells, from which neutrophils were notably abundant across the samples. A subset of basal keratinocytes in the HRASG12V condition exhibited a transcriptional program enriched for tumour-propagating markers lamc2 and cd44a, alongside genes upregulated in proliferation and metabolism. This preneoplastic (PNC) state was detected in both biological replicates at 24 hpi and was followed by a proliferative but lamc2- cd44a- state at 48 hpi, suggesting loss of cancer stem-like cell population. Comparison with hierarchical clustering showed agreement in the identification of the PNC population. Neutrophils were consistently enriched in the HRASG12V condition, indicating an associated immune response. These results demonstrate that Stator is able to resolve cell identity for the early detection of oncogenic reprogramming in a vertebrate model, providing one framework for investigating early tumorigenesis at single-cell resolution. Read More
Related posts:
Design and Implementation of Artificial Intelligence Algorithms for Prognostic Stratification in Acute Myeloid Leukemia
Advances in Flexible and Lightweight III-V Multijunction Solar Cells for High Power Densities
Asymmetric Cell-DEVS models with the Cadmium simulator
Prediction of cell types and subtypes underlying nervous system phenotypes from snRNAseq data
Manufacturing process for III-V multijunction solar cells on germanium substrates with a total thickness below 60 microns
Evaluation and Comparison of Spatial Clustering for Solar Irradiance Time Series