Recognizing Splicing Junction Acceptors in Eukaryotic Genes Using Hidden Markov Models and Machine Learning Methods

Document Type

Conference Proceeding

Publication Date

12-1-2000

Abstract

The development of the hidden Markov model (HMM) acceptor model for splicing junction acceptor sites recognition was discussed. An HMM with 16 states and a set of transitions was defined for modeling a true acceptor site. The states and transitions were represented as a digraph where states corresponded to vertices and transitions to edges. Each state was associated with a discrete output probability distribution. The performance evaluation of the HMM system for true acceptor sites showed that on average, the system correctly detected 91.9% of the true acceptor sites in the test data.

Identifier

0006555749 (Scopus)

ISBN

[0964345692]

Publication Title

Proceedings of the Joint Conference on Information Sciences

First Page

786

Last Page

789

Issue

2

Volume

5

This document is currently not available here.

Share

COinS