Ured that only high-quality reads were used for subsequent analysis (Additional file 2: Figure S2). Uniquely mapped reads were assigned to a genome locus based on the highest probability. Multimapping reads were assigned using criteria specific to each pipeline (see Methods). Reads with no match were not used for locus assignment. The three pipelines yielded similar numbers of total mapped reads (Additional file 3: Table S1). The correlations among all resulting mappings were examined by pair-wise comparisons. The resulting R2 values showed a high correlation among libraries from the same genetic background (Additional file 4: Figure S3). The assembly files were then filtered by same criteria, fold change 2 and false discovery rate (FDR) < 0.1, to yield SR45 differentially regulated (SDR) genes. A total of 739, 760 and 391 SR45-upregulated genes, and a total of 1052, 921 and 805 SR45-downregulated genes were identified by Tophat2, STAR and LG12, respectively (Additional file 5:Table S2). After comparing the identity of these genes from all three pipelines, a total of 89 PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/25746230 common genes were determined to be upregulated by SR45, and a total of 269 common genes were determined to be downregulated by SR45 (Fig. 2a). While many positive gene candidates were no doubt excluded by applying such a highly stringent filter, combining the three different pipelines reduced the chance of false positives due to pipelinespecific bias during the assembly process, and provides us with a robust set of common genes that are up-regulated or down-regulated by SR45 with high confidence. Therefore, we followed up with these common genes in the subsequent analysis. No significant enrichment of gene ontology (GO) categories was observed for SR45-upregulated genes (those with reduced expression in the sr45? mutant), which suggests that SR45 promotes the expression of genes in diverse pathways. However, several GO categories were significantly enriched with high confidence among SR45-downregulated genes (those with elevated expression in the sr45? mutant) (Fig. 2b). When we compared the GO enrichment pattern in all analyses (Tophat2, STAR, LG12) individually (Additional file 6: Table S3), we found that all of the GO terms enriched in the common gene set showed similar or much better enrichment than was seen in the three individual pipelines. This suggests that the common gene set does indeed contain reduced background noise. However, the p-values for each GO category in the common gene set were slightly get Roc-A higher in most categories than in the individual pipelines due to a smaller number of the remaining genes in the common gene set. The 89 SR45-upregulated genes are involved in diverse biological processes, from metabolism (Glucose-6-phosphate 1-dehydrogenase 4, isocitrate lyase, and GDSL esterase), transport (peptide transporter 4, sodium/metabolite cotransporter BASS5 and BASS6), auxin biosynthesis (YUC4), kinases (MAPKKK21 and GWD2) to transcriptional/post-transcriptional regulation (RDR3, SR45, CID9) (Additional file 5: Table S2). Selected genes were confirmed by qPCR (Fig. 2c). Despite the floral phenotype of sr45?, none of these genes are known to directly or specifically contribute to flower or seed development. However, we think it is likely that most, if not all, of these SR45-upregulated genes are not primary targets of SR45. Since SR45 has been recognized as a splicing activator, it is plausible that SR45-dependent alternative splicing more directly acts on.