Fig. 6
From: Dictionary learning allows model-free pseudotime estimation of transcriptomic data
3D visualisation of the coefficients for a dictionary with three atoms for two datasets. Shown are three rotations of the 3D visualisation. Visualisation for dataset GSE84712 are shown on the left; Visualisations for dataset E-MTAB-6811 are shown on the right. For each dataset, the low-dimensional representation is coloured once by the subtype separating feature (Dose for GSE84712 and Organ for E-MTAB-6811) and once by the experimental times. For dataset GSE84712, samples with a dose of 30 μM are clearly separated from those with no or 3 μM lead exposure for higher times. On each branch, samples are well ordered according to experimental times. A similar pattern is observable for dataset E-MTAB-6811