Optimizing the Cell Painting assay for image-based profiling

Extended Data Fig. 1 The JUMP-MOA plate map labeled by compound.

The JUMP-MOA compound plate map: unlabeled wells are DMSO only; all other wells are labeled to show the distribution of compound replicates across the entire plate. A version of this figure grouped by MOA rather than compound is available as Fig. 2a .

Source data

Extended Data Fig. 2 Further assessment of cell type suitability for Cell Painting.

a , Percent replicating of each cell line and time point tested when selecting the cell type and treatement time points to be used by the consortium. A549 (red dots) has better replication at both time points of compound treatment than U2OS (blue dots) but worse performance at all time points of genetic perturbation experiments. b , Assessment of decreasing or increasing the number of cells plated by 20% on the ability to perform percent matching. Few consistent effects are seen when comparing across the columns present in each panel. c , Assessment of using parental cell lines or Cas9-expressing polyclonal lines for compound treatments. Few consistent effects are seen; the top right panel represents the desired task, the ability to match compounds added to either parental or Cas9 cells to CRISPR (which must be performed in Cas9 cells). The results in that panel do not seem extremely different either for matching 48-h compound treatment to 96 h of CRISPR treatment (compare the orange dots) or for matching 48-h compound treatment to 144 h of CRISPR treatment (compare the yellow dots). For more information including the plate(s) represented by each data point, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 .

Source data

Extended Data Fig. 3 Assessment of the effect of gene-treatment-related compounds on Cell Painting performance.

a , Addition of blasticidin to ORF plates or puromycin to CRISPR plates does not appear to improve cross-modailty matching across modalities versus unselected plates; compare the second and third columns to the first column in the top center, top-left and bottom-middle panels. b , Addition of selection compounds may have a deleterious effect on percent replicating versus unselected plates, although we cannot rule out that this is due to fewer replicates for the selected plates than the unselected ones. c , Addition of 4 µg/ml polybrene for 24 h may produce a phenotypic effect; polybrene addition displays decreased inter-treatment cross-plate percent replicating versus intra-treatment cross-plate-percent replicating (compare the center column to the outside columns), even though both sets of plates were stained and imaged as part of the same batch. d , Addition of polybrene to Target2-treated cells does not improve percent matching between Target2-treated plates and ORF-treated plates; note that the Target-ORF plates came from a previous batch and were not stained and imaged in parallel to the Target2 compound plates. For more information including the plate(s) represented by each data point, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 .

Source data

Extended Data Fig. 4 Assessment of permeabilization timing and plate type on Cell Painting performance.

a , Performing permeabilization at the same time as staining produces comparable percent replicating and percent matching results. b , PhenoPlates without barrier wells and Aurora plates with barrier wells produce comparable percent replicating and percent matching results. For more information including the plate(s) represented by each data point, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 .

Source data

Extended Data Fig. 5 Expanded assessment of dye concentration effects on Cell Painting performance.

a , Within batches, reducing all dyes by two- or fourfold produces comparable percent replicating and comparable but perhaps slightly decreased percent matching results. b , All quantitatively tested stain concentration changes, broken out by the dye(s) perturbed. For more information including the plate(s) represented by each data point, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 .

Source data

Extended Data Fig. 6 Further assessment of imaging conditions on Cell Painting performance.

a , Imaging of the same plates on a wide-field microscope with 2 × 2 binning versus a different manufacturer’s microscope in confocal mode with 1 × 1 binning. No major differences are observed. b , Acquisition of multiple Z planes slightly improves percent replicating but not percent matching in two batches. For more information including the plate(s) represented by each data point, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 .

Source data

Extended Data Fig. 7 Assesment of Cell Painting performance when using various combinations of channel features.

Mean percent replicating of eight JUMP-MOA plates stained with the final staining conditions after dropping out all feature names containing individual channel names before performing feature selection and calculation of percent replicating. ‘None’ means that only AreaShape features are present. To create a sufficiently compact data representation, the eight channels present were split four each onto the X (AGP, DNA, ER and Mito) and Y (RNA, Brightfield, BFHigh and BFLow) axes; this allows visualization of the 256 possible unique combinations. Note that these results are not the same as truly having the stains not present, because (i) a channel still may have been used in creating the initial segmentation results, and (ii) it does not account for cross-talk between stains. An alternate representation of these data is presented in Extended Data Fig. 8 . For more information including the plates represented in this figure, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 . AGP, actin golgi plasma membrane; BF, brightfield; Mito, mitochondria.

Extended Data Fig. 8 Change in Cell Painting performance when adding individual channel features.

Change in mean percent replicating of eight JUMP-MOA plates stained with the final staining conditions when an individual stain is present versus not; each chart shows for a particular stain when added to the non-channel-specific features plus zero or more other stains (x axis) the change in the mean percent replicating (y axis) when those features are included. The color(s) of each marker indicate which channel(s) are already present. Note that these results are not the same as truly having the stains not present, because (i) a channel still may have been used in creating the initial segmentation results, and (ii) it does not account for cross-talk between stains. The absolute percent replicating numbers are available in Extended Data Fig. 7 . For more information including the plates represented in this figure, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data File 1 .

Extended Data Fig. 9 Assessment of Cell Painting performance for individual channels and feature categories.

Stain-by-stain breakdown of mean percent replicating of eight JUMP-MOA plates stained with the final staining conditions after dropping out all possible combinations of features from the five stain-specific feature categories before performing feature selection and calculation of percent replicating. Unlike the analyses in Extended Data Fig. 7 and Extended Data Fig. 8 , non-channel-specific feature categories (AreaShape and Neighbors) are not included here, because the goal is to assess the information contribution of each feature category in each stain. To create sufficiently compact data representation, the five categories present were split three each onto the X (Correlation, Granularity and Intensity) and two onto the Y (RadialDistribution and Texture) axes; this allows visualization of the 31 possible unique combinations. For more information including the plates represented in this figure, see the Source Data file for this figure; expanded experimental details for each plate may be found in Supplementary Data 1 .

Source data

Supplementary Information

Supplementary Table 1, Methods and References

Supplementary Data 1

Treatment, staining and microscopy conditions of all plates included in this work