Limits...
NAViGaTing the micronome--using multiple microRNA prediction databases to identify signalling pathway-associated microRNAs.

Shirdel EA, Xie W, Mak TW, Jurisica I - PLoS ONE (2011)

Bottom Line: Since microRNA binding is sequence-based but possibly structure-specific, work in this area has resulted in multiple databases storing predicted microRNA:target relationships computed using diverse algorithms.Our pathway-based analysis of mirDIP data suggests microRNAs are involved in intra-pathway signalling.We identify two distinct classes of microRNAs, suggesting a hierarchical organization of microRNAs co-targeting genes both within and between pathways, and implying differential involvement of universe and intra-pathway microRNAs at the disease level.

View Article: PubMed Central - PubMed

Affiliation: Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada.

ABSTRACT

Background: MicroRNAs are a class of small RNAs known to regulate gene expression at the transcript level, the protein level, or both. Since microRNA binding is sequence-based but possibly structure-specific, work in this area has resulted in multiple databases storing predicted microRNA:target relationships computed using diverse algorithms. We integrate prediction databases, compare predictions to in vitro data, and use cross-database predictions to model the microRNA:transcript interactome--referred to as the micronome--to study microRNA involvement in well-known signalling pathways as well as associations with disease. We make this data freely available with a flexible user interface as our microRNA Data Integration Portal--mirDIP (http://ophid.utoronto.ca/mirDIP).

Results: mirDIP integrates prediction databases to elucidate accurate microRNA:target relationships. Using NAViGaTOR to produce interaction networks implicating microRNAs in literature-based, KEGG-based and Reactome-based pathways, we find these signalling pathway networks have significantly more microRNA involvement compared to chance (p<0.05), suggesting microRNAs co-target many genes in a given pathway. Further examination of the micronome shows two distinct classes of microRNAs; universe microRNAs, which are involved in many signalling pathways; and intra-pathway microRNAs, which target multiple genes within one signalling pathway. We find universe microRNAs to have more targets (p<0.0001), to be more studied (p<0.0002), and to have higher degree in the KEGG cancer pathway (p<0.0001), compared to intra-pathway microRNAs.

Conclusions: Our pathway-based analysis of mirDIP data suggests microRNAs are involved in intra-pathway signalling. We identify two distinct classes of microRNAs, suggesting a hierarchical organization of microRNAs co-targeting genes both within and between pathways, and implying differential involvement of universe and intra-pathway microRNAs at the disease level.

Show MeSH

Related in: MedlinePlus

MicroRNA interaction network characteristics.Examination of four microRNA interaction network characteristics across well-known signalling pathways using KEGG (panels A and B) and Reactome pathway databases (panels C and D). Signalling pathways tend to be enriched for the number of microRNAs, the number interactions and the number of high degree nodes mapped.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3045450&req=5

pone-0017429-g005: MicroRNA interaction network characteristics.Examination of four microRNA interaction network characteristics across well-known signalling pathways using KEGG (panels A and B) and Reactome pathway databases (panels C and D). Signalling pathways tend to be enriched for the number of microRNAs, the number interactions and the number of high degree nodes mapped.

Mentions: After initially testing our hypothesis on the PI3 Kinase pathway, we decided to undertake a more extensive and rigorous examination of signalling pathways within the cell. Since pathway definitions have not been unanimously settled and there is still much debate as to which resource defines a signalling pathway most accurately and comprehensively, we decided to use pathways delineated by the Kyoto Encyclopedia of Genes and Genomes database (KEGG) [70], [71] and pathways defined by the Reactome [72], [73], [74] database to further support the microRNA networks built based on expert-curated pathway reviews in the previous section. Examining interactions predicted at 2 threshold levels: 2+DB and 3+DB, we created microRNA networks for both the canonical signalling pathways and for 2000 permutations of pathways created with the same number of primary node genes. Our findings showed a similar trend for most interaction sets and signalling pathways that we examined. We found that true signalling pathways tend to involve more microRNAs and contain more interactions, as well as having more high degree nodes (degree ≥4) than pathways created out of a random set of starting nodes. We examined 9 KEGG pathways and 12 Reactome pathways at the 2+DB and 3+DB interaction thresholds. The pathways with the lowest average p-values (that is the average of p-values across the 4 measured parameters - number of network interactions, number of network microRNAs, number of network nodes with degree ≥4 and network density) were KEGG pathways: ERBB signalling pathway (hsa04012) (2+DB), mTOR signalling pathway (hsa04150) (2+DB), Wnt signalling pathway (hsa04310) (2+DB), MAPK signalling pathway (hsa04010) (3+DB) and Pathways in cancer (hsa05200) (3+DB) with average p-values of p<0.0006, p<0.0009, p<0.002, p<0.002, p<0.007, respectively (Figure 5). Of the pathways described in both the KEGG and Reactome databases (NOTCH, VEGF and WNT), WNT results were the least conserved across both databases – showing significance in KEGG (average p-values of p<0.002 and p<0.036 for 2+DB and 3+DB respectively), but not in Reactome (average p-values of p<0.64 and p<0.68 for 2+DB and 3+DB respectively), while NOTCH measured parameters were the most likely to be consistent across the two databases (average p-values of p<0.102 and p<0.105 for 2+DB and 3+DB respectively in KEGG and average p-values of p<0.256 and p<0.139 for 2+DB and 3+DB respectively in Reactome). We found that some pathways had greater tendencies than others to show significance – for example the FGFR and Cell Cycle Genes pathways (which, it could be argued, is not a signalling pathway and hence does not fit within this study and hence acts as our negative control) described only by the Reactome database had a tendency towards higher p-values than other pathways examined (Reactome FGFR pathway average p-values of p<0.35 and p<0.4 for 2+DB and 3+DB respectively and Reactome Cell Cycle Genes average p-values of p<0.78 and p<0.45 for 2+DB and 3+DB respectively). The measured parameters found to be most frequently significant across all studied scenarios were the number of microRNA nodes in the network with degree ≥4 (significant at p<0.05 in 30/42 tested scenarios), and the number of total microRNA:target interactions in the network (significant at p<0.05 in 27/42 tested scenarios). As highlighted in Figure 5 – one can find enrichments that are supported by both pathway databases, while other enrichments are highlighted in the analysis using one or the other pathway database. Examining expert-curated pathways, KEGG pathways and Reactome pathways with similar findings gives us confidence that this phenomenon is in fact real.


NAViGaTing the micronome--using multiple microRNA prediction databases to identify signalling pathway-associated microRNAs.

Shirdel EA, Xie W, Mak TW, Jurisica I - PLoS ONE (2011)

MicroRNA interaction network characteristics.Examination of four microRNA interaction network characteristics across well-known signalling pathways using KEGG (panels A and B) and Reactome pathway databases (panels C and D). Signalling pathways tend to be enriched for the number of microRNAs, the number interactions and the number of high degree nodes mapped.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3045450&req=5

pone-0017429-g005: MicroRNA interaction network characteristics.Examination of four microRNA interaction network characteristics across well-known signalling pathways using KEGG (panels A and B) and Reactome pathway databases (panels C and D). Signalling pathways tend to be enriched for the number of microRNAs, the number interactions and the number of high degree nodes mapped.
Mentions: After initially testing our hypothesis on the PI3 Kinase pathway, we decided to undertake a more extensive and rigorous examination of signalling pathways within the cell. Since pathway definitions have not been unanimously settled and there is still much debate as to which resource defines a signalling pathway most accurately and comprehensively, we decided to use pathways delineated by the Kyoto Encyclopedia of Genes and Genomes database (KEGG) [70], [71] and pathways defined by the Reactome [72], [73], [74] database to further support the microRNA networks built based on expert-curated pathway reviews in the previous section. Examining interactions predicted at 2 threshold levels: 2+DB and 3+DB, we created microRNA networks for both the canonical signalling pathways and for 2000 permutations of pathways created with the same number of primary node genes. Our findings showed a similar trend for most interaction sets and signalling pathways that we examined. We found that true signalling pathways tend to involve more microRNAs and contain more interactions, as well as having more high degree nodes (degree ≥4) than pathways created out of a random set of starting nodes. We examined 9 KEGG pathways and 12 Reactome pathways at the 2+DB and 3+DB interaction thresholds. The pathways with the lowest average p-values (that is the average of p-values across the 4 measured parameters - number of network interactions, number of network microRNAs, number of network nodes with degree ≥4 and network density) were KEGG pathways: ERBB signalling pathway (hsa04012) (2+DB), mTOR signalling pathway (hsa04150) (2+DB), Wnt signalling pathway (hsa04310) (2+DB), MAPK signalling pathway (hsa04010) (3+DB) and Pathways in cancer (hsa05200) (3+DB) with average p-values of p<0.0006, p<0.0009, p<0.002, p<0.002, p<0.007, respectively (Figure 5). Of the pathways described in both the KEGG and Reactome databases (NOTCH, VEGF and WNT), WNT results were the least conserved across both databases – showing significance in KEGG (average p-values of p<0.002 and p<0.036 for 2+DB and 3+DB respectively), but not in Reactome (average p-values of p<0.64 and p<0.68 for 2+DB and 3+DB respectively), while NOTCH measured parameters were the most likely to be consistent across the two databases (average p-values of p<0.102 and p<0.105 for 2+DB and 3+DB respectively in KEGG and average p-values of p<0.256 and p<0.139 for 2+DB and 3+DB respectively in Reactome). We found that some pathways had greater tendencies than others to show significance – for example the FGFR and Cell Cycle Genes pathways (which, it could be argued, is not a signalling pathway and hence does not fit within this study and hence acts as our negative control) described only by the Reactome database had a tendency towards higher p-values than other pathways examined (Reactome FGFR pathway average p-values of p<0.35 and p<0.4 for 2+DB and 3+DB respectively and Reactome Cell Cycle Genes average p-values of p<0.78 and p<0.45 for 2+DB and 3+DB respectively). The measured parameters found to be most frequently significant across all studied scenarios were the number of microRNA nodes in the network with degree ≥4 (significant at p<0.05 in 30/42 tested scenarios), and the number of total microRNA:target interactions in the network (significant at p<0.05 in 27/42 tested scenarios). As highlighted in Figure 5 – one can find enrichments that are supported by both pathway databases, while other enrichments are highlighted in the analysis using one or the other pathway database. Examining expert-curated pathways, KEGG pathways and Reactome pathways with similar findings gives us confidence that this phenomenon is in fact real.

Bottom Line: Since microRNA binding is sequence-based but possibly structure-specific, work in this area has resulted in multiple databases storing predicted microRNA:target relationships computed using diverse algorithms.Our pathway-based analysis of mirDIP data suggests microRNAs are involved in intra-pathway signalling.We identify two distinct classes of microRNAs, suggesting a hierarchical organization of microRNAs co-targeting genes both within and between pathways, and implying differential involvement of universe and intra-pathway microRNAs at the disease level.

View Article: PubMed Central - PubMed

Affiliation: Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada.

ABSTRACT

Background: MicroRNAs are a class of small RNAs known to regulate gene expression at the transcript level, the protein level, or both. Since microRNA binding is sequence-based but possibly structure-specific, work in this area has resulted in multiple databases storing predicted microRNA:target relationships computed using diverse algorithms. We integrate prediction databases, compare predictions to in vitro data, and use cross-database predictions to model the microRNA:transcript interactome--referred to as the micronome--to study microRNA involvement in well-known signalling pathways as well as associations with disease. We make this data freely available with a flexible user interface as our microRNA Data Integration Portal--mirDIP (http://ophid.utoronto.ca/mirDIP).

Results: mirDIP integrates prediction databases to elucidate accurate microRNA:target relationships. Using NAViGaTOR to produce interaction networks implicating microRNAs in literature-based, KEGG-based and Reactome-based pathways, we find these signalling pathway networks have significantly more microRNA involvement compared to chance (p<0.05), suggesting microRNAs co-target many genes in a given pathway. Further examination of the micronome shows two distinct classes of microRNAs; universe microRNAs, which are involved in many signalling pathways; and intra-pathway microRNAs, which target multiple genes within one signalling pathway. We find universe microRNAs to have more targets (p<0.0001), to be more studied (p<0.0002), and to have higher degree in the KEGG cancer pathway (p<0.0001), compared to intra-pathway microRNAs.

Conclusions: Our pathway-based analysis of mirDIP data suggests microRNAs are involved in intra-pathway signalling. We identify two distinct classes of microRNAs, suggesting a hierarchical organization of microRNAs co-targeting genes both within and between pathways, and implying differential involvement of universe and intra-pathway microRNAs at the disease level.

Show MeSH
Related in: MedlinePlus