Integrating large-scale functional genomic data has significantly accelerated our understanding of gene functions. However, no algorithm has been developed to differentiate functions for isoforms of the same gene using high-throughput genomic data. This is because standard supervised learning requires ‘ground-truth’ functional annotations, which are lacking at the isoform level. To address this challenge, we developed a generic framework that interrogates public RNA-seq data at the transcript level to differentiate functions for alternatively spliced isoforms. For a specific function, our algorithm identifies the ‘responsible’ isoform(s) of a gene and generates classifying models at the isoform level instead of at the gene level. Through cross-validation, we demonstrated that our algorithm is effective in assigning functions to genes, especially the ones with multiple isoforms, and robust to gene expression levels and removal of homologous gene pairs. We identified genes in the mouse whose isoforms are predicted to have disparate functionalities and experimentally validated the ‘responsible’ isoforms using data from mammary tissue. With protein structure modeling and experimental evidence, we further validated the predicted isoform functional differences for the genes Cdkn2a and Anxa6. Our generic framework is the first to predict and differentiate functions for alternatively spliced isoforms, instead of genes, using genomic data. It is extendable to any base machine learner and other species with alternatively spliced isoforms, and shifts the current gene-centered function prediction to isoform-level predictions.

The vast majority of multi-exon genes in humans undergo alternative splicing, which greatly increases the functional diversity of protein species. Predicting functions at the isoform level is essential to further our understanding of developmental abnormalities and cancers, which frequently exhibit aberrant splicing and dysregulation of isoform expression. However, determination of isoform function is very difficult, and efforts to predict isoform function have been limited in the functional genomics field. Now deep sequencing of RNA provides an unprecedented amount of expression data at the transcript level. We describe here emerging computational approaches that integrate such large-scale RNA-seq data for predicting functions of alternatively spliced isoforms, and we discuss their applications in developmental and cancer biology. We outline future directions for isoform function prediction, emphasizing the need for heterogeneous genomic data integration and tissue-specific, dynamic isoform-level network modeling, which will allow the field to realize its full potential.

Canonical isoforms in different databases have been defined as the most prevalent, most conserved, most expressed, longest, or the one with the clearest description of domains or post-translational modifications. In this article, we revisit these definitions of canonical isoforms based on functional genomics and proteomics evidence, focusing on mouse data. We report a novel functional relationship network-based approach for identifying the Highest Connected Isoforms (HCIs). We show that 46% of these HCIs are not the longest transcripts. In addition, this approach revealed many genes that have more than one highly connected isoforms. Averaged across 175 RNA-seq datasets covering diverse tissues and conditions, 65% of the HCIs show higher expression levels than non-highest connected isoforms (NCIs) at the transcript level. At the protein level, these HCIs highly overlap with the expressed splice variants, based on proteomic data from eight different normal tissues. These results suggest that a more confident definition of canonical isoforms can be made through integration of multiple lines of evidence, including highest connected isoforms defined by biological processes and pathways, expression prevalence at the transcript level, and relative or absolute abundance at the protein level. This integrative proteogenomics approach can successfully identify principal isoforms that are responsible for the canonical functions of genes.

Objective To elucidate the molecular mechanisms involved in renal inflammation during the progression, remission and relapse of nephritis in murine lupus models using transcriptome analysis. Methods Kidneys from NZB/W F1 and NZM2410 mice were harvested at intervals during their disease course or after remission induction. Genome wide expression profiles were obtained from microarray analysis of perfused kidneys. Real time PCR analysis for selected genes was used to validate the microarray data. Comparisons between groups using SAM, and unbiased analysis of the entire dataset using singular value decomposition and self-organizing map were performed. Results Few changes in the renal molecular profile were detected in pre-nephritic kidneys but a significant shift in gene expression, reflecting inflammatory cell infiltration and complement activation occurred at proteinuria onset. Subsequent changes in gene expression predominantly affected mitochondrial dysfunction and metabolic stress pathways. Endothelial cell activation, tissue remodeling and tubular damage were the major pathways associated with loss of renal function. Remission induction reversed most, but not all of the inflammatory changes and progression towards relapse was associated with recurrence of inflammation, mitochondrial dysfunction and metabolic stress signatures. Conclusion Immune cell infiltration and activation is associated with proteinuria onset and reverses with immunosuppressive therapy but disease progression is associated with renal hypoxia and metabolic stress. Optimal therapy of SLE nephritis may therefore need to target both immune and non-immune disease mechanisms. In addition, the overlap of a substantial subset of molecular markers with those expressed in human lupus kidneys suggests potential new biomarkers and therapeutic targets.