1
Supplementary Figure 1 Analyses of RNAi phenotypes in Kamath et al., Gonczy et al. and Maeda et al. datasets. a. Comparison between RNAi phenotypes and published loss-of-function phenotypes of known loci. Shown are the numbers of loci with Nonv, Vpep or any phenotype (All) that were screened in each study (‘Scrn’), the number that were detected in each screen (‘Hit’), and the percentage detected. To determine the rate of false positives (‘False +ve’) in each screen, we examined known loci with published non-lethal loss-of-function phenotypes that had been analysed in each screen — any of these having a Nonv phenotype by RNAi were classed as false positives (‘Hit’). b. Comparison of RNAi phenotypes in Kamath et al., Gonczy et al. and Maeda et al. datasets. The total number of genes screened in common in any pair of screens is shown in plain type beside the central double-headed arrows; also shown is the total number of genes that gave a Nonv phenotype when pooling both screens. In the overlapping circles, we show the percentages of these Nonv genes that were found in common between any pair of screens or uniquely in one or other screen; numbers of genes in each category are shown on parentheses. The high percentage of Nonv phenotypes found in the Maeda et al dataset but absent from Kamath et al or Gonczy et al can be partially explained by a higher rate of false positive Nonv RNAi phenotypes in the Maeda et al screen.
Supplementary Figure 2. Relative representation of genes with different molecular functions on each chromosome. Genes were placed into one of the following putative molecular functional classes as described in Methods: protein synthesis (P Synth), RNA synthesis (RNA Synth), DNA synthesis and repair/cell cycle (DNA/CC), cellular architecture (Cell Arch), RNA binding (RNA Bind), chromatin regulation (Chromatin), protein degradation (Degrad), energy and intermediary metabolism (Metab), transcription factors (Txn Factor), nucleic-acid binding (NA Bind), signal transduction (Signaling), small molecule transport (SM Txpt), specific proteases (Protease), retroviral- and transposon-derived sequences (Viral), collagens (Collagen), and genes with neuronal functions (Neuro); genes which could not be assigned to a functional class were placed into the Unknown class (Unknown). Shown are the relative frequencies of occurrence of genes in each functional class for each chromosome. Bars in yellow denote a statistically significant overenrichment (p<0.01); red bars represent an underenrichment (p<0.01).
Supplementary Figure 3. Regions of chromosomal enrichment (blue) and underenrichment (green) of microarray mounts 0-29 from Kim et al (2001) Science293, 2087-92. Yellow and teal boxes represent the arm and cluster regions, respectively of the six C. elegans chromosomes. Black lines inside these show regions enriched for duplications. Shown above each chromosome in one line for each mount are chromosomal regions significantly (p<0.01) enriched (blue) or underenriched (green) in a 250 gene window. Roman numerals label the chromosomes. Above and to the left of each chromosome, numbers refer to the mount plotted. Note that mount 2 is composed almost entirely of the genes identified in Blumenthal et al.40 to be transpliced to SL2, and thus it marks downstream genes in operons (T. Blumenthal and S. Kim, personal communication). The regions enriched for mount 2 overlap nearly identically with the regions enriched for Nonv genes (red bars in Figure 4b).