Amyotrophic lateral sclerosis (ALS) is a rapidly fatal neurodegenerative disorder classically characterised by motor neuron death. However, genetic and histopathological overlap with frontotemporal dementia (FTD) suggests extra-motor involvement, and that pathology likely extends to other neuronal populations (Chapter 1). This may be especially evident in patients with the hexanucleotide repeat expansion in the gene C9ORF72 (c9HRE): the most common genetic cause of both ALS and FTD. We hypothesise that in c9HRE-related neurodegeneration, neuronal pathology may extend beyond the motor neurons. To date, a comprehensive survey of selective neuronal vulnerability has not yet been performed. With recent advances in single-cell and spatial transcriptomic approaches, our hypothesis can now be investigated in an unbiased fashion and at unprecedented resolution. The work in this thesis focuses on identifying selectively vulnerable neurons in the post-mortem primary motor cortex (M1) of patients with ALS and FTD caused by a . M1 tissue was processed for single nuclei RNA-sequencing (snRNA-seq) and 10X Visium spatial transcriptomics (Chapter 2). I first investigated cortical neuron vulnerability using snRNA-seq (Chapter 3). Upper motor neurons (herein called L5 ETs: layer 5 extratelencephalic-projecting excitatory neurons) were expected to display pathology in c9HRE and served as a positive control. Other vulnerable neuronal populations were identified by examining neurons with similar transcriptomic responses to L5 ETs. In so doing, upper layer excitatory subtypes, PVALB+ fast-spiking basket interneurons, and specific VIP+ interneurons were identified as vulnerable. Transcriptomic changes primarily manifested in neurons as altered mitochondrial, proteostatic, and synaptic function. Minimal evidence of glial involvement was observed. Given the range of developmental origins amongst vulnerable populations and the relative sparing of layer 6, I hypothesised that microcircuit connectivity might determine vulnerability. I investigated this hypothesis using Visium (Chapter 4). Three lines of evidence support the L6 sparing previously inferred from snRNA-seq. First, inferred vulnerable populations are located outside L6. Second, L6 transcriptomic changes are enriched for axon-related genes, and may thus originate from neurons with soma outside L6. Third, signature deconvolution suggests expected pathology is significantly higher for c9HRE samples for layers superficial to L5. Taken together, my spatial transcriptomic data supports L6 sparing inferred from snRNA-seq, which in turn reinforces the notion of microcircuitry being predictive of vulnerability (rather than spatial proximity or developmental origin). I also cross-examined the biological pathways highlighted in the snRNA-seq analysis with findings from literature (Chapter 5). Identified pathways were preserved upon metaanalysis, and suggest a role for mTORC signalling. Importantly, I also observed that existing c9HRE model systems may not be appropriate for understanding post-mortem changes. In summary, this work suggests that selective vulnerability in c9HRE is determined by (and possibly spreads through) neuronal connections. Deeper examination of this hypothesis is limited by existing technology, but nascent approaches that merge connectomics with transcriptomics (Chapter 6) may eventually allow us to investigate this in future.