Teton™ CytoProfiling, Correlating Teton with RNA-Seq (HeLa, HUVEC, A549 Cells)

Background

Teton™ CytoProfiling data generated from three different cell lines was benchmarked against bulk RNA-seq data to assess its technical and biological concordance with this standard gene expression method. Teton CytoProfiling on AVITI24™ expands beyond traditional bulk RNA-seq by providing multimodal cellular context alongside subcellular spatial RNA and protein data. Learn more about this dataset in our Correlating Teton with RNA-Seq Technical Note.

Methods

HeLa, HUVEC, and A549 cell lines were grown and maintained under standard growth conditions. For Teton CytoProfiling, cells were seeded onto PLL-coated slides (one 12-well slide for each cell line) and flow cells were prepared according to the Teton CytoProfiling User Guide. Each flow cell was run on AVITI24 using the Teton Human MAPK Cell Cycle Panel kit, which includes 350 RNA targets, 50 protein and phospho-protein targets, and 6 cell paint markers.

Bulk RNA expression was also performed on AVITI24. RNA was first extracted from 500,000 cells from each cell line with the Direct-zol RNA kit (Zymo Research). RNA libraries were generated with the RNA Library Prep Kit with Polaris Depletion (Watchmaker), then pooled and sequenced on AVITI24 with a 2x75 Cloudbreak Freestyle™ High Output Sequencing Kit.

Teton data was filtered and adjusted by batch with the CytoProfiling package from the Element CytoProfiling Repository. Teton RNA targets were matched with the bulk RNA-seq targets to create a gene list that only included the RNA targets shared between both datasets. The Teton data was further adjusted by probe concentration and normalized using counts per 10,000 cells. RNA-seq data was processed and normalized using transcripts per million (TPM) counts. Teton data was pseudobulked by well to assess well-to-well reproducibility. To facilitate the comparison with the bulk-RNA-seq data, Teton data from individual cells was pseudobulked by cell line.

Pearson correlations between the Teton CytoProfiling runs and bulk RNA-seq data were performed for each cell line after the data from both assays was normalized and log-transformed. Differential expression concordance between Teton and bulk RNA-seq was also evaluated across pairwise comparisons of the HeLa, A549, and HUVEC cell lines. Concordance for these pairwise comparisons was assessed by computing the Pearson correlation between the matched differences in log-transformed expression values across shared genes and by calculating directional agreement.

Additional methods details and results are provided in the Correlating Teton with RNA-seq Technical Note.

Results

Teton CytoProfiling provides RNA and differential expression data that correlates well with bulk RNA-seq, enabling simple benchmarking to validate assay results. In a comparison between the Teton and bulk RNA-seq runs, Pearson correlation coefficients were 0.835 for HeLa, 0.836 for A549, and 0.722 for HUVEC. Differential expression analysis also demonstrated concordance between Teton and bulk RNA-seq across pairwise comparisons of the cell lines, with 97.3% directional agreement between Teton and RNA-seq in the HeLa vs A549 comparison, 84% directional agreement in the HUVEC vs HeLa comparison, and 85.3% directional agreement in the HUVEC vs A549 comparison.

Register below to download the data and analysis notebook used to generate the correlations presented in the Tech Note.