Skip to main content
. 2024 May 7;15:3840. doi: 10.1038/s41467-024-48041-6

Fig. 3. tauFisher accurately predicts circadian time when the training and test data are from different assay methods.

Fig. 3

tauFisher trained on mouse skin microarray data can predict circadian time for skin bulk RNAseq data. a Overlay of the predictor gene expression in GSE38622 (training) and GSE83855 (test). b Prediction outcomes from tauFisher. tauFisher trained on mouse SCN bulk RNAseq data can predict circadian time of pseudobulk data generated from mouse SCN scRNAseq data. c Overlay of the predictor gene expression in GSE157077 (training) and GSE117295 (test). d Prediction outcomes from tauFisher. e Prediction outcomes from TimeSignatR and ZeitZeiger when trained on mouse SCN bulk RNAseq and tested on mouse SCN scRNAseq pseudobulk. tauFisher trained on mouse skin microarray data can predict circadian time of pseudobulk data generated from dermis scRNAseq data. f Overlay of the predictor gene expression in GSE38622 (training) and GSE223109 (test). g Prediction outcomes from tauFisher. b, d, e, g The dashed lines mark where predictions equal truth. RMSE: root mean square error. Source data are provided as a Source Data file.