Supplemental Data for:
Lewis et al., Cell 115, pp. 787–798

Figure S1A. The locations and secondary structures of the TargetScan predicted miRNA binding sites are diagrammed for the 115 predicted vertebrate targets (representing 107 distinct genes) conserved between human, mouse, rat, and Fugu. For each conserved miRNA-UTR pair, the name of the miRNA, the names of the orthologous target mRNAs in each species, and the Ensembl description for the human gene are shown. The next line contains the length of the sequence used as the human UTR and the positions of the 5’ ends of the seed matches, followed by schematic of the UTR where each dash represents 25 nt and the location of each seed match is represented by a “1.” Below the UTR schematic are the TargetScan generated duplex structures used to score the UTR. Each sequence consists of the miRNA sequence joined to the binding site by a “&,” with the predicted base pairing indicated by the parentheses below the sequence. Calculated free energies of binding for the duplexes (in kcal/mol) are shown next to the structures. Schematics, structures, and energies are shown for all orthologous genes considered by TargetScan in predicting these targets.

Figure S1B. The locations and secondary structures of the TargetScan predicted miRNA binding sites are diagrammed for the 854 predicted mammalian targets (representing 442 distinct genes) of rMamm miRNAs conserved between human, mouse, and rat. The format is similar to that of Figure S1A.

Figure S2. The locations and secondary structures of predicted miRNA binding sites are diagramed for the 13 tested human genes and their mammalian orthologs, as well as the lin-41 control, using the format of Figure S1A. The portion of the human UTR that was PCR amplified and cloned downstream of the luciferase reporter is highlighted in blue in the UTR schematic. The sequences of the wild type and mutant PCR products are indicated at the end of each record. Seed matches are shown in blue for the wild-type sequence and the corresponding sequences are highlighted in red for the mutant sequence, with mutated residues shown in lowercase. The last record shows sequences of the lin-41 controls (lin-41, a known target of C. elegans let-7 RNA) referred to in Figure 3, with regions of complementarity to let-7 RNA underlined and mutations shown in red lowercase letters.

Table S1. The Human miRNAs and Shuffled Controls used in This Study
The inclusion of each miRNA in the three subsets used in this study (rMamm, nrMamm, and nrVert) is indicated by Y (Yes) or N (No). For those miRNAs in nrMamm, the sequences of the four shuffled variants generated my MiRhuffle are listed on the next four lines (labeled miR-X_sh0, miR-X_sh1, etc.). The 19 miRNAs not in rMamm are those for which Fugu homologs could not be identified. For completeness, we list here a few additional human miRNAs that were not used in this study for various reasons, such as imperfect conservation between human and mouse: miR-17, miR-103b, miR-129, miR-134, miR-150, miR-151, miR-189 miR-200a, miR-217.

Table S2. The Predicted Mammalian Targets for miRNAs in rMamm
The 442 genes in this set were predicted as targets of rMamm miRNAs by TargetScan in human, mouse, and rat orthologs, as described in the text. MiRNAs with identical seeds that are predicted to target the same gene are shown in a single row of the table. MiRNAs with different seeds that target the same gene are listed on separate lines.

Table S3. Targets of Shuffled Control Sequences
The 558 shuffled sequence:UTR pairs found human, mouse, and rat that were predicted by TargetScan for any of the four cohorts of MiRshuffled variants of nrMamm miRNAs are shown.

Table S4. Biological Function Classes of Predicted miRNA Targets and Controls
The number and percentage of UTRs annotated in various Gene Ontology biological process categories are shown for the 400 predicted miRNA-UTR pairs predicted by TargetScan for nrMamm miRNAs; the miRNA-UTR pairs predicted with randomized miRNAs (average of 4 cohorts); and for the total set of orthologous genes conserved between human, mouse, and rat. For cases in which GO categories with a parent-child relationship are shown, the child is indented. Note that the GO categories are not mutually exclusive.

Table S5. The Predicted Vertebrate Targets for miRNAs in nrVert
The orthologous genes for this set scored highly as targets of nrVert miRNAs in all four organisms studied. MicroRNAs with different seeds that target the same gene are listed on separate lines.