This gene was likely identified by the genetic algorithm as a pu tative regulat

Additional file 1 shows the target genes of the PCa associated DE miRNAs. At last, 1236, 3566, 1520 and 4749 target genes of the DE miRNAs of four different datasets were obtained respectively. The identification of the microRNA regulatory pathways in prostate cancer [You must be registered and logged in to see this link.] The collection of the four different datasets are from different platforms, the overlapping of miRNA probes between these data are about 40 60% while the detected differently expressed miRNA profiles only have 3% over lapping. We aim to identify the consistent pattern at high level. First, the target genes of DE miRNAs found by at least 3 datasets were extracted, then mapped to function and pathway databases, e. g. GO, KEGG and GeneGO, to identify PCa associated functions and pathways.

In this process, we identified 1221 target genes of the PCa associated DE miRNAs, among which 253 were shared by all the four target gene datasets, and 968 overlapped in three of the four datasets. As shown in Figure 2, the ligand indepen dent activation of ESR1 and ESR2 is the most [You must be registered and logged in to see this link.] significant GeneGO pathway. In Figure 2, insulin like growth factor 1 encodes the protein involved in mediating growth and development. In this pathway, IGF 1 binds to IGF 1 receptor on the membrane and activates signal transduction through Shc, SOS, Mek1, and ERK2, finally mediating the production of ESR1 and ESR2. Genes involved in the signal transduction above are all target genes of highly expressed miRNAs in pros tate cancer samples, therefore, the expression of ESR1 and ESR2 will be down regulated which is in accordance with the previous report by Gamba and his co authors.

Figure 3 illustrates various biological themes enriched in the gene list. The [You must be registered and logged in to see this link.] left side of the figure is a bar plot of enriched GO terms, KEGG pathways, and GeneGO pathways against log10, the top five terms of each biological theme were shown in the right. The details are also available in Additional files 3, 4, and 5. In these files, the pathway or GO terms were sorted by p value. Overall, we identified 38 GO terms, 16 KEGG pathways, and 99 GeneGO pathways that are enriched with target genes of the PCa associated DE miRNAs. Analysis and validation of the putative microRNA regulatory pathways in prostate cancer Among the 99 enriched GeneGO pathways, 67 pathways were also significantly enriched in our pre vious study in which we processed 10 mRNA microarray datasets.

In the set of top 15 GeneGO pathways in our previous work, 11 were also detected in the 99 pathways in this study. To identify potential microRNA regulatory pathways in prostate cancer, the 15 most significantly enriched pathways were chosen for the analysis. Of those, four had previously been reported to be related to prostate cancer in PubMed citations. We verified the other 11 pathways indirectly by analysis of the component genes in PubMed citations although the wet lab experiments can direct validate them. Among the top 15 pathways reported by both the previously and the present studies, 3 pathways are the same in both studies, 2 of the 3 pathways are novel ones i. e, 1 ligand independent activation of ESR1 and ESR2, this is the most significant pathway we mentioned in the last section, and 2 membrane bound ESR1, inter action with growth factors signalling.