Figure 2: Analysis of DDA samples.

a–f, Similarity of DDA samples at the protein level, analysed by principal component analysis (PCA) and multidimensional scaling (MDS). a, PCA analysis of all 32 samples analysed by DDA mass spectrometry, each with 5000 cells. Red ovals highlight samples prepared in batch #2. The percentage of the variance contributed by each principal component is indicated in the axis. b, PCA analysis of the 24 vetted DDA samples. Note that the samples cluster together more tightly than in A. c, PCA analysis of the 24 vetted DDA samples, reanalysed with the requirement of two or more unique peptides supporting each identification. d, MDS analysis of all 32 samples analysed by DDA mass spectrometry. e, MDS analysis of the 24 vetted DDA samples. f, MDS analysis of the 24 vetted DDA samples, reanalysed with the requirement of two or more unique peptides supporting each identification. Note the close correspondence of each pair of replicates, with the exception of one of the P7 utricle GFP-negative replicates (blue arrow in panels D-F). The largest variance is associated with GFP state and the second variance component separates samples by developmental time. Legend in F applies to all panels. g,h, Distribution of riBAQ values for proteins in each DDA sample. Boxes contain the upper (UQ) and lower quartiles (LQ), with the median value indicated by a horizontal line and the interquartile distance (IQD) being UQ-LQ. The remainder of the data are within the lines, with the exception of outliers (defined as >UQ+1.5•IQD or <LQ-1.5•IQD), plotted as individual points. The numbers on the top indicate the total number of proteins or protein groups detected in each sample; the value is coloured red if <1000.