This website contains supplementary material used in SpliceIT. SpliceIT (Splice Identification Technique) is a hybrid method for splice site prediction consisting of:

a) a Gaussian SVM classification using probabilistic modeling not previously used, and

b) a binary decision tree trained on a small set of known splicing elements.

Background: Splice sites define the boundaries of exonic regions and dictate protein synthesis and function. The post-transcriptional mechanism undertaking splice site selection involves complex interactions among positional and compositional features of different lengths. Computational modeling of the underlying constructive information is especially challenging in order to decipher splicing-inducing elements and alternative splicing factors.

Results: SpliceIT introduces a hybrid method for splice site prediction that employs probabilistic and biology-driven modeling in two subsequent classification steps. The first step is undertaken by a Gaussian SVM that is trained on two novel position-dependent feature encodings. The extracted predictions are then combined with known species-specific regulatory elements in order to induce a tree-based modeling. The overall classification performance on human and Arabidopsis thaliana splice datasets shows that SpliceIT outperforms current state-of-the-art predictors. This is achieved without compromising space complexity and in a time-effective way.

Conclusions: We show that by modeling complementary evidence, probabilistic and biology-driven, in subsequent classification steps it is feasible to increase splice signal identification and to give grounds for further investigation of the underlying biological factors presiding over constitutive and alternative splicing.

REFERENCE
SpliceIT: A hybrid method for splice signal identification based on probabilistic and biological inference. Malousi A, Chouvarda I, Koutkias V, Kouidou S, Maglaveras N. Journal of Biomedical Informatics (in press)

 
     

Lab. of Medical Informatics, Medical School
Aristotle University of Thessaloniki