Analyze data
The tools on this page evaluate your own predictions in the browser with RSC, normalized pairwise accuracy and the pairwise difference loss of the paper. Computed results opens files written by the toolkit, and Conventional metrics computes whole-section correlation and AUROC for comparison with RSC.
Score tables and bundles written by the toolkit open here.
Examples are the score table behind Figure 6c,d (11 models, nine Xenium samples of breast, pancreatic and lung cancer) and a full analysis of one lung cancer sample.
Drop your file hereor use the file picker
Files are processed in your browser and stay on your computer. Data privacy
Conventional metrics computes whole-section correlation, correlation within regions and in larger bins, and AUROC against expert annotations for comparison with RSC. In the paper, whole-section correlation overstates prediction quality within tissue regions, and Pearson r with measured RNA expression understates prediction quality at fine scales (Fig. 1b,c of the paper).
Loading…