Table App_1: All Available datasets of protein-protein interactions (IntAct) Yeast-2-Hybrid experiments are highlighted in blue. DatasetOrganismMethod (only large-scale)# Interactions arahy small scalepeanut15 arath small scalemouse-ear cress772 bacan small scaleb. anthracis5 bovin small scalecow9 caeel Li 2004c. elegansYeast-2-Hybrid4073 caeel small scalec. elegans673 canfa small scaledog1 chick small scalechicken15 dicdi small scaleslime mold1 drome Giot 2003fruit flyYeast-2-Hybrid20667 drome Stanyon 2004fruit flyYeast-2-Hybrid1615 drome Formstecher 2005fruit flyYeast-2-Hybrid1674 drome small scalefruit fly878 ecoli small scalee. coli2290 helpy small scaleh. pylori12 horvu small scalebarley139 human Bouwmeester 2004humanTandem Affinity Purification1670 human small scalehuman33841 mesau small scalegolden hamster1 mouse small scalemouse5495 pethy small scalepetunia111 rabit small scalerabbit1 rat small scalerat807 rhopa small scaler. palustris1398 salti small scales. typhi1 schpo small scalefission yeast38 sheep small scalesheep1 shifl small scales. flexneri4 theth small scalet. thermophilus1 trybr small scalet. brucei rhodesiense59 xenla small scaleAfrican clawed frog29 yeast Uetz 2000baker's yeastYeast-2-Hybrid1495 yeast Ito 2001baker's yeastYeast-2-Hybrid4524 yeast Gavin 2002baker's yeastTandem Affinity Purification19254 yeast Ho 2002baker's yeastMass Spectrometry29098 yeast small scalebaker's yeast1632 Figure App_1: Each curve above shows the accuracy (red) as shown in Figure 3 and the number of true positives counted at a certain HSSP-value cutoff (green). Figure App_2 (A-D): A: Different sampling of intra- vs. inter-species: we allowed transfers of the type A-B to A'-B or A-B to A-B' (Methods). The performance became significantly better for intra-species PPI-transfers, thus further widening the gap between intra and inter-species transfers. B: Inclusion of transfers within the same data set: we included homology transfers within the same experimental dataset (Methods). The effect was very similar to those observed for different sampling (#1), i.e. widening the gap between intra- and inter-species inferences. C: Using TAP-like data (Table App_1, Supplementary Online Material) as a constraint for the negatives. To illustrate this assume that TAP pulled down a complex of six proteins. While we cannot infer that all 15 possible interactions are physical, all could be. Therefore, we ignored a false postive prediction (did not count it) if we could find the interaction in those 15 TAP protein-protein pairs. The accuracy slightly increased for both yeast vs yeast (intra-species) comparisons as well as for non-yeast vs yeast (inter-species) comparisons. Note that yeast is the only organism with available TAP-like data. D: We used a redundant dataset (instead of a non-redundant, bias-reduced set) from organism o (Fig. 7) to hunt for interologs in organism p (Fig. 7). The main message indicated by the results for this latter experiment (#4) stays the same as in our original procedure (Methods): Intra species comparisons are more accurate than inter-species comparisons. Due to more samples in the dataset for organism o (Fig. 7) and thus higher counts, the errors slightly decreased.
