In silico prediction of splice-altering single nucleotide variants in the human genome

In silico tools have been developed to predict variants that may have an impact on pre-mRNA splicing. The major limitation of the application of these tools to basic research and clinical practice is the difficulty in interpreting the output. Most tools only predict potential splice sites given a DNA sequence without measuring splicing signal changes caused by a variant. Another limitation is the lack of large-scale evaluation studies of these tools. We compared eight in silico tools on 2959 single nucleotide variants within splicing consensus regions (scSNVs) using receiver operating characteristic analysis. The Position Weight Matrix model and MaxEntScan outperformed other methods. Two ensemble learning methods, adaptive boosting and random forests, were used to construct models that take advantage of individual methods. Both models further improved prediction, with outputs of directly interpretable prediction scores. We applied our ensemble scores to scSNVs from the Catalogue of Somatic Mutations in Cancer database. Analysis showed that predicted splice-altering scSNVs are enriched in recurrent scSNVs and known cancer genes. We pre-computed our ensemble scores for all potential scSNVs across the human genome, providing a whole genome level resource for identifying splice-altering scSNVs discovered from large-scale sequencing studies.     

In vitro and in silico analysis reveals an efficient algorithm to predict the splicing consequences of mutations at the 5' splice sites

We have found that two previously reported exonic mutations in the PINK1 and PARK7 genes affect pre-mRNA splicing. To develop an algorithm to predict underestimated splicing consequences of exonic mutations at the 5′ splice site, we constructed and analyzed 31 minigenes carrying exonic splicing mutations and their derivatives. We also examined 189 249 U2-dependent 5′ splice sites of the entire human genome and found that a new variable, the SD-Score, which represents a common logarithm of the frequency of a specific 5′ splice site, efficiently predicts the splicing consequences of these minigenes. We also employed the information contents (R_i) to improve the prediction accuracy. We validated our algorithm by analyzing 32 additional minigenes as well as 179 previously reported splicing mutations. The SD-Score algorithm predicted aberrant splicings in 198 of 204 sites (sensitivity = 97.1%) and normal splicings in 36 of 38 sites (specificity = 94.7%). Simulation of all possible exonic mutations at positions −3, −2 and −1 of the 189 249 sites predicts that 37.8, 88.8 and 96.8% of these mutations would affect pre-mRNA splicing, respectively. We propose that the SD-Score algorithm is a practical tool to predict splicing consequences of mutations affecting the 5′ splice site.     

Redirecting splicing with bifunctional oligonucleotides

Ectopic modulators of alternative splicing are important tools to study the function of splice variants and for correcting mis-splicing events that cause human diseases. Such modulators can be bifunctional oligonucleotides made of an antisense portion that determines target specificity, and a non-hybridizing tail that recruits proteins or RNA/protein complexes that affect splice site selection (TOSS and TOES, respectively, for targeted oligonucleotide silencer of splicing and targeted oligonucleotide enhancer of splicing). The use of TOSS and TOES has been restricted to a handful of targets. To generalize the applicability and demonstrate the robustness of TOSS, we have tested this approach on more than 50 alternative splicing events. Moreover, we have developed an algorithm that can design active TOSS with a success rate of 80%. To produce bifunctional oligonucleotides capable of stimulating splicing, we built on the observation that binding sites for TDP-43 can stimulate splicing and improve U1 snRNP binding when inserted downstream from 5′ splice sites. A TOES designed to recruit TDP-43 improved exon 7 inclusion in SMN2. Overall, our study shows that bifunctional oligonucleotides can redirect splicing on a variety of genes, justifying their inclusion in the molecular arsenal that aims to alter the production of splice variants.     

Characterization and prognostic significance of alternative splicing events in lower‐grade diffuse gliomas

Alternative splicing (AS) is assumed to play important roles in the progression and prognosis of cancer. Currently, the comprehensive analysis and clinical relevance of AS in lower‐grade diffuse gliomas have not been systematically addressed. Here, we gathered alternative splicing data of lower‐grade diffuse gliomas from SpliceSeq. Based on the Percent Spliced In (PSI) values of 515 lower‐grade diffuse glioma patients from the Cancer Genome Atlas (TCGA), we performed subtype‐differential AS analysis and consensus clustering to determine robust clusters of patients. A total of 48 050 AS events in 10 787 genes in lower‐grade diffuse gliomas were profiled. Subtype‐differential splicing analysis and functional annotation revealed that spliced genes were significantly enriched in numerous cancer‐related biological phenotypes and signalling pathways. Consensus clustering using AS events identified three robust clusters of patients with distinguished pathological and prognostic features. Moreover, each cluster was also associated with distinct genomic alterations. Finally, we developed and validated an AS‐related signature with Cox proportional hazards model. The signature, significantly associated with clinical and molecular features, could serve as an independent prognostic factor for lower‐grade diffuse gliomas. Thus, our results indicated that AS events could discriminate molecular subtypes and have prognostic impact in lower‐grade diffuse gliomas.     

Human Splicing Finder: an online bioinformatics tool to predict splicing signals

Thousands of mutations are identified yearly. Although many directly affect protein expression, an increasing proportion of mutations is now believed to influence mRNA splicing. They mostly affect existing splice sites, but synonymous, non-synonymous or nonsense mutations can also create or disrupt splice sites or auxiliary cis-splicing sequences. To facilitate the analysis of the different mutations, we designed Human Splicing Finder (HSF), a tool to predict the effects of mutations on splicing signals or to identify splicing motifs in any human sequence. It contains all available matrices for auxiliary sequence prediction as well as new ones for binding sites of the 9G8 and Tra2-β Serine-Arginine proteins and the hnRNP A1 ribonucleoprotein. We also developed new Position Weight Matrices to assess the strength of 5′ and 3′ splice sites and branch points. We evaluated HSF efficiency using a set of 83 intronic and 35 exonic mutations known to result in splicing defects. We showed that the mutation effect was correctly predicted in almost all cases. HSF could thus represent a valuable resource for research, diagnostic and therapeutic (e.g. therapeutic exon skipping) purposes as well as for global studies, such as the GEN2PHEN European Project or the Human Variome Project.     

Predicting human splicing branchpoints by combining sequence-derived features and multi-label learning methods

Background

Alternative splicing is the critical process in a single gene coding, which removes introns and joins exons, and splicing branchpoints are indicators for the alternative splicing. Wet experiments have identified a great number of human splicing branchpoints, but many branchpoints are still unknown. In order to guide wet experiments, we develop computational methods to predict human splicing branchpoints.

Results

Considering the fact that an intron may have multiple branchpoints, we transform the branchpoint prediction as the multi-label learning problem, and attempt to predict branchpoint sites from intron sequences. First, we investigate a variety of intron sequence-derived features, such as sparse profile, dinucleotide profile, position weight matrix profile, Markov motif profile and polypyrimidine tract profile. Second, we consider several multi-label learning methods: partial least squares regression, canonical correlation analysis and regularized canonical correlation analysis, and use them as the basic classification engines. Third, we propose two ensemble learning schemes which integrate different features and different classifiers to build ensemble learning systems for the branchpoint prediction. One is the genetic algorithm-based weighted average ensemble method; the other is the logistic regression-based ensemble method.

Conclusions

In the computational experiments, two ensemble learning methods outperform benchmark branchpoint prediction methods, and can produce high-accuracy results on the benchmark dataset.

     

DNA methylation profiling to assess pathogenicity of BRCA1 unclassified variants in breast cancer

Germline pathogenic mutations in BRCA1 increase risk of developing breast cancer. Screening for mutations in BRCA1 frequently identifies sequence variants of unknown pathogenicity and recent work has aimed to develop methods for determining pathogenicity. We previously observed that tumor DNA methylation can differentiate BRCA1-mutated from BRCA1-wild type tumors. We hypothesized that we could predict pathogenicity of variants based on DNA methylation profiles of tumors that had arisen in carriers of unclassified variants. We selected 150 FFPE breast tumor DNA samples [47 BRCA1 pathogenic mutation carriers, 65 BRCAx (BRCA1-wild type), 38 BRCA1 test variants] and analyzed a subset (n=54) using the Illumina 450K methylation platform, using the remaining samples for bisulphite pyrosequencing validation. Three validated markers (BACH2, C8orf31, and LOC654342) were combined with sequence bioinformatics in a model to predict pathogenicity of 27 variants (independent test set).  Predictions were compared with standard multifactorial likelihood analysis. Prediction was consistent for c.5194-12G>A (IVS 19-12 G>A) (P>0.99); 13 variants were considered not pathogenic or likely not pathogenic using both approaches. We conclude that tumor DNA methylation data alone has potential to be used in prediction of BRCA1 variant pathogenicity but is not independent of estrogen receptor status and grade, which are used in current multifactorial models to predict pathogenicity.     

Genome wide identification and classification of alternative splicing based on EST data

MOTIVATION: Alternative splicing is currently seen to explain the vast disparity between the number of predicted genes in the human genome and the highly diverse proteome. The mapping of expressed sequences tag (EST) consensus sequences derived from the GeneNest database onto the genome provides an efficient way of predicting exon-intron boundaries, gene structure and alternative splicing events. However, the alternative splicing events are obscured by a large number of putatively artificial exon boundaries arising due to genomic contamination or alignment errors. The current work describes a methodology to associate quality values to the predicted exon-intron boundaries. High quality exon-intron boundaries are used to predict constitutive and alternative splicing ranked by confidence values, aiming to facilitate large-scale analysis of alternative splicing and splicing in general. RESULTS: Applying the current methodology, constitutive splicing is observed in 33,270 EST clusters, out of which 45% are alternatively spliced. The classification derived from the computed confidence values for 17 of these splice events frequently correlate (15/17) with RT-PCR experiments performed for 40 different tissue samples. As an application of the confidence measure, an evaluation of distribution of alternative splicing revealed that majority of variants correspond to the coding regions of the genes. However, still a significant fraction maps to non-coding regions, thereby indicating a functional relevance of alternative splicing in untranslated regions. AVAILABILITY: The predicted alternative splice variants are visualized in the SpliceNest database at http://splicenest.molgen.mpg.de     

Alternative splicing and protein structure evolution

Alternative splicing is thought to be one of the major sources for functional diversity in higher eukaryotes. Interestingly, when mapping splicing events onto protein structures, about half of the events affect structured and even highly conserved regions i.e. are non-trivial on the structure level. This has led to the controversial hypothesis that such splice variants result in nonsense-mediated mRNA decay or non-functional, unstructured proteins, which do not contribute to the functional diversity of an organism. Here we show in a comprehensive study on alternative splicing that proteins appear to be much more tolerant to structural deletions, insertions and replacements than previously thought. We find literature evidence that such non-trivial splicing isoforms exhibit different functional properties compared to their native counterparts and allow for interesting regulatory patterns on the protein network level. We provide examples that splicing events may represent transitions between different folds in the protein sequence–structure space and explain these links by a common genetic mechanism. Taken together, those findings hint to a more prominent role of splicing in protein structure evolution and to a different view of phenotypic plasticity of protein structures.     

Temperature-dependent splicing of beta-globin pre-mRNA

A T→G mutation at nucleotide 705 of human β-globin intron 2 creates an aberrant 5′ splice site and activates a cryptic 3′ splice site upstream. In consequence, the pre-mRNA is spliced via aberrant splice sites, despite the presence of the still functional correct sites. Surprisingly, when IVS2-705 HeLa or K562 cells were cultured at temperatures below 30°C, aberrant splicing was inhibited and correct splicing was restored. Similar temperature effects were seen for another β-globin pre-mRNA, IVS2-745, and in a construct in which a β-globin intron was inserted into a coding sequence of EGFP. Temperature-induced alternative splicing was affected by the nature of the internal aberrant splice sites flanking the correct sites and by exonic sequences. The results indicate that in the context of thalassemic splicing mutations and possibly in other alternatively spliced pre-mRNAs, temperature is one of the parameters that affect splice site selection.