Fig. 2From: Optimal dimensionality selection for independent component analysis of transcriptomic dataClassification of independent components across dimensions for each dataset. The PC-VA, OptICA, and MSTD dimensions are shown where applicable. The four datasets are A PRECISE 1.0, B PRECISE 2.0, C B. subtilis, and D StaphPRECISE. Note that since no single gene components are found for the StaphPRECISE dataset, the non-single gene and robust component lines overlapBack to article page