macrohet: single-cell trajectory dataset of intracellular Mycobacterium tuberculosis population heterogeneity in iPSC-derived macrophages

This dataset contains the processed single-cell trajectory data underlying the single-cell analyses in: Day et al., "Fast-growing intracellular Mycobacterium tuberculosis populations evade antibiotic treatment", Nature Communications (2026). The data were generated from long-term live-cell timelapse microscopy of induced pluripotent stem cell-derived macrophages (iPSDMs) infected with fluorescently labelled Mycobacterium tuberculosis and imaged across 6 drug conditions. Individual macrophages were segmented and tracked over time, and per-cell, per-timepoint measurements were extracted, yielding >20,000 single-cell trajectories across 121 fields of view and 4 biological replicates. The data are provided as a single Apache Parquet file (single_cell_trajectories.parquet) with one row per cell per timepoint. Columns include: # identifiers & indexing 'Unique ID', 'ID', 'Cell ID', 'Acquisition ID', 'Experiment ID', 'Frame', 'Time (hours)', # well / plate position 'row', 'col', # experimental conditions 'Strain', 'Compound', 'Concentration', 'Condition', 'Biological Replicate', 'Technical Replicate', # spatial 'x', 'y', 'x_um', 'y_um', 'MSD', # Mtb burden 'Mtb Area (µm)', 'dMtb Area (µm)', 'dMtb Area between frames (µm)', 'Mtb Area Processed (µm)', # fluorescence intensities 'GFP', 'RFP', # host macrophage morphology 'Mphi Area (µm)', 'dMphi Area (µm)', 'Eccentricity', # infection state 'Infection Status', 'Initial Infection Status', 'Final Infection Status', # track QC (mtb_origin corresponds to Uptake/Transfer/Intrinsic Growth'Edge Status', 'mtb_origin', # growth model outputs 'Time Model (hours)', 'Mtb Area Model (µm)', 'r2', 'Doubling Times', 'Doubling Amounts', 'Doubling Frames', The analysis code used to generate this dataset and to reproduce the manuscript figures is available at https://github.com/nthndy/macrohet and archived on Zenodo (DOI: 10.5281/zenodo.20819399). The repository includes a small, downscaled example acquisition sufficient to run the pipeline end to end. Licensed under CC-BY 4.0. If you use these data, please cite the associated publication and this record.