ASPIC: a web resource for alternative splicing prediction and transcript isoforms characterization

Abstract

Alternative splicing (AS) is now emerging as a major mechanism contributing to the expansion of the transcriptome and proteome complexity of multicellular organisms. The fact that a single gene locus may give rise to multiple mRNAs and protein isoforms, showing both major and subtle structural variations, is an exceptionally versatile tool in the optimization of the coding capacity of the eukaryotic genome. The huge and continuously increasing number of genome and transcript sequences provides an essential information source for the computational detection of genes AS pattern. However, much of this information is not optimally or comprehensively used in gene annotation by current genome annotation pipelines. We present here a web resource implementing the ASPIC algorithm which we developed previously for the investigation of AS of user submitted genes, based on comparative analysis of available transcript and genome data from a variety of species. The ASPIC web resource provides graphical and tabular views of the splicing patterns of all full-length mRNA isoforms compatible with the detected splice sites of genes under investigation as well as relevant structural and functional annotation. The ASPIC web resource—available at —is dynamically interconnected with the Ensembl and Unigene databases and also implements an upload facility.

What is a gene? An updated operational definition

“…2 shows the splicing pattern of the gene CDKN2A, as determined by the ASPIC program (Castrignano et al, 2006). It should be noted that the first and second transcripts (CDKN2A.Ref and CDKN2A.Tr2 in Fig.…”

PIntron: A fast method for gene structure prediction via maximal pairings of a pattern and a text

“…Other comparisons among gene-structure prediction tools in literature [4] are performed over different datasets, therefore that analysis and ours are not completely comparable. Anyway, for illustrative purposes we report that two commonly used software packages for predicting the exon-intron structure, namely ECgene [16] and ASPic [8], have sensitivity/specificity that is respectively 0.92/0.63 and 0.88/0.77 [4]: not as good as those obtained by our pipeline. Another parameter that is surely interesting is the number of EST supporting a exon/intron.…”

“…Indeed, it must be pointed out that few efforts have been done in the direction of providing a formal framework to design efficient algorithms for the general AS prediction problem. As a matter of fact, existing tools are not able to process efficiently genes that are huge or with a very large set of associated clusters of ESTs [8]. A basic reason of this fact is that combinatorial methods for the problem must combine two different steps: (1) to produce putative spliced alignments of ESTs against the gene region and (2) to use redundancy and the whole covering of the gene region provided by a cluster in the prediction process by selecting among putative spliced alignments of multiple ESTs [2] the ones that confirm the same gene structure.…”

GenePC AND ASPIC INTEGRATE GENE PREDICTIONS WITH EXPRESSED SEQUENCE ALIGNMENTS TO PREDICT ALTERNATIVE TRANSCRIPTS

“…ASPIC then performs a multiple genome-EST alignment optimized for producing a set of transcripts with a minimal number of exons. 11,14 The longest open reading frame is then determined for each transcript output by ASPIC and overlapping CDS spans are assigned to different compatibility groups or bins. The number of bins is determined by the locus with the largest number of alternative transcripts.…”