Skip to main content
. 2004 Feb 17;101(9):2846–2851. doi: 10.1073/pnas.0306638101

Fig. 4.

Fig. 4.

(a) Distribution of CD in eukaryotic innovation domains (squares) and prokaryotic-only domains (circles). The data for these histograms are binned into bins of size 0.2. The CD is calculated as explained in Methods. (b) Distribution of TrC8 normalized by domain length in prokaryotic-only Dali domains (circles) and eukaryotic innovation domains (squares). The data were binned with bin size 4,000. Only domains with CD in the range between 3 and 4 are taken; 262 eukaryotic innovation domains and 843 prokaryotic-only domains fall in this range. The Kolmogorov-Smirnov (KS) P value for the null hypothesis that these two datasets were drawn from the same distribution is 0.001. For control, we randomly split prokaryotic-only domains ensemble into two equal parts and compared their distributions of TrC8. In contrast to comparison between eukaryotic innovation and prokaryotic-only domains, these two sets appear to be identical: KS P value for the same null hypothesis is 0.647.