O'Hora, B.; Laddach, R.; Nuamah, R.; Alberts, E.; Wall, I.; Bell, J.; Johnston, D. A.; James, S.; Norman, J.; Jones, M. G.; Chiappini, C.; Grigoriadis, A.; Quist, J.

Motivation: Advances in spatial transcriptomics technologies at single-cell resolution have highlighted the need for innovative quality assessment approaches and improved analytical tools. Imaging-based spatial transcriptomics technologies, such as the CosMx Spatial Molecular Imager (SMI), provide the location and abundance of transcripts through multifocal imaging. Optical sections (or Z-slices) form a Z-stack that represents the tissue depth. Transcript dispersion can be observed across these Z-slice and introduce considerable levels of technical noise to the data that can negatively impact downstream analysis. Package Functionality: NoButter is an R package designed to evaluate transcript dispersion in CosMx SMI spatial transcriptomics data. Using the raw data, the transcript distribution is as-sessed for each Z-slice of a Z-stack across multiple fields of views (FOVs). To sys-tematically identify transcript dispersion, the percentage of transcripts located out-side cell boundaries is calculated. Z-slices exhibiting high levels of transcript dis-persion can be excluded, while high-confidence transcripts are preserved. Usage Scenario: To demonstrate the functionalities of NoButter, spatial transcriptomics data was generated using the CosMx SMI for lymph node tissue, a lung sample, and two tri-ple-negative breast cancers (TNBCs). Use cases illustrate substantial transcript dispersion in optical planes closer to the glass slide. In these Z-slices, on average, an additional 10% of the transcripts were discarded using NoButter. Cleaning such Z-slices with high dispersion rates reduces technical noise and improves the over-all quality of the spatial transcriptomics data. Availability: The package can be accessed at https://github.com/cancerbioinformatics/NoButter.