Reproducing "Method of moments framework for differential expression analysis of single-cell RNA sequencing data"

Input data to reproduce the figures of Kim et al., Cell 2024, using the memento analysis code at 10.5281/zenodo.13637731. Each archive holds one figure's inputs, organized by panel. Download only the figure you need — the archives are independent. FileFigureDownloadUnpacked figure2_data.tar.gzFigure 2 — method validation and comparisons1.3 GB4.1 GB figure3_data.tar.gzFigure 3 — interferon stimulation in human airway epithelium2.7 GB9.6 GB figure4_data.tar.gzFigure 4 — Perturb-seq of transcription factor knockouts1.6 GB4.9 GB figure5_data.tar.gzFigure 5 — eQTL, vQTL and coexpression QTL analysis5.8 GB21 GB figure6_data.tar.gzFigure 6 — memento in CZI CELLxGENE Discover44 MB44 MB Unpack an archive and point MEMENTO_DATA_PATH at the directory containing it, then run that figure's script: tar -xzf figure3_data.tar.gz export MEMENTO_DATA_PATH=$PWD cd publication/figure3 && python make_figure3.py Verify each download against its .sha256 file before unpacking. Full instructions, and what each figure should produce, are in publication/. Three figures need something beyond their archive: Figure 3, panel F additionally needs supplementary Table S1E of Mostafavi et al., Cell 2016 (mmc2.xls). That is a publisher supplementary file and is not redistributed here; the figure README gives the DOI to fetch it from. Figure 5, panels F-I additionally need individual-level genotypes, which are controlled-access under dbGaP phs002812.v1.p1 and are not included. Panels A-E are complete: the minor-allele-frequency filter panel A applies ships as an aggregate per-variant frequency, which contains no individual-level data. Figure 6, panels C and D additionally query the public CELLxGENE census at run time, so that figure needs network access.