Legend to Figure S1
Figure title: Protein identification by TAP/MS
In order to assess the reliability of the employed protein identification strategy, we compared the distribution of protein size and isoelectric point of our data (A) to the respective distribution as calculated for ORFs predicted from the yeast genome (B). The superimposition of the two sets (C) shows a very high degree of overlap indicating that there appears to be no general bias towards protein size or isoelectric point. Identified proteins cover sizes from 6.6 kDa to 559.3 kDa and the pI range between 3.9 and 12.4. By plotting the ratio of the number of proteins identified in this study and the number of all yeast ORFs against protein size (D), it appears that small proteins (<15 kDa) are underrepresented in this study. This observation might reflect difficulties associated with unreliable migration behavior of small proteins in 1D PAGE, prediction of small ORFs from genomic sequences and the fact that protein identification by peptide mass fingerprinting is hampered by the small number of tryptic peptides that are often available from digests of small proteins.
3