Spinal cord injury triggers complex cellular and molecular responses that differ substantially between species. While laboratory mice have limited regenerative capacity after central nervous system injury, the spiny mouse (Acomys) has emerged as a mammalian model with unusually strong tissue repair and regenerative features. This project aims to compare the transcriptional programs activated after spinal cord injury in Acomys and conventional laboratory mouse, with the long-term goal of identifying conserved and species-specific mechanisms associated with neural damage response, inflammation, glial activation, neuronal stress, and potential regenerative competence. The project will use high-throughput transcriptomic and multi-omic datasets generated from injured and control spinal cord tissue and/or dissociated neural cell populations. We will perform cross-species integration of gene expression profiles, ortholog mapping, cell-type annotation, differential expression analysis, pathway enrichment, and trajectory-based modelling of injury-induced cellular states. Particular attention will be given to damage-responsive neuronal programs, including AP-1-associated stress genes such as Atf3, regeneration-associated genes, inflammatory signaling modules, and glia–neuron interaction signatures. By comparing the timing, magnitude, and cell-type specificity of these transcriptional responses between Acomys and mouse, the project seeks to define molecular programs that may distinguish a more regeneration-permissive injury response from a more scar- and inflammation-dominated response. A central computational component of the project is the use of deep learning and large-scale statistical modelling for cross-species representation learning. These methods will be used to align homologous cell states across species, learn latent transcriptional programs after injury, infer gene regulatory modules, and identify candidate molecular features predictive of regenerative or non-regenerative responses. Because the datasets involve high-dimensional single-cell or spatially resolved omics profiles, repeated model training, parameter optimization, and large-scale matrix operations, substantial computational resources are required. Access to dedicated compute infrastructure will enable efficient preprocessing, model development, benchmarking, and reproducible downstream analysis. The project is supervised by Dr. Enric Llorens, CMB, KI. The proposed work will contribute to a mechanistic understanding of mammalian spinal cord injury responses and may help identify candidate pathways for future experimental validation in neural repair and regenerative medicine.