Dear colleagues,
I would like to invite you to the lecture "From Cell Painting Quality Control to Compound Similarity: An Interactive R/Shiny Application" presented by Anna Šišková. The workshop will be on Wednesday October 7, 2026 at 9AM in room the 1.09 at IMTM.
Abstract
The Cell Painting assay generates rich morphological profiles, and careful quality control is essential before any biological interpretation. An R/Shiny application guides researchers through the full analysis workflow, from raw profiles to mechanism-of-action exploration.
Quality control includes well-level cell count filtering, plate-level inspection via per-channel intensity maps, Moran's I across all extracted features to flag spatial artifacts (with follow-up visualization on plate maps), and pairwise cosine similarity of positive control wells.
Once data passes QC, an interactive UMAP projection reveals batch effects across plates; batch correction is currently under development. Replicate reproducibility is then quantified per compound by the median pairwise cosine similarity of its replicates, with an adjustable threshold determining which compounds carry forward. Only compounds meeting that threshold proceed to the final stage: comparison of median morphological profiles within each dataset independently and between user-defined test and reference sets.