# Difference Analysis Among Six Kinds of Acceptor Splicing Sequences by the Dispersion Features of 6-mer Subsets in Human Genes

## TL;DR

This study uses dispersion features of 6-mer subsets to identify differences in human gene acceptor splicing sequences across six splicing modes.

## Contribution

The study introduces dispersion features of XY1 6-mer subsets to distinguish splicing modes and sub-modes in human genes.

## Key findings

- Dispersion features effectively distinguish three splicing modes and three sub-modes of alternative splicing.
- Significant differences in dispersion distributions were observed across upstream, core, and downstream regions of splicing sites.
- XY1 6-mer subset dispersion values reveal compositional patterns and base correlations in acceptor splicing sequences.

## Abstract

We first constructed a comprehensive dataset of human gene acceptor splicing sequences, which were classified into common splicing, constitutive splicing, and alternative splicing modes. Alternative splicing was further subdivided into normal, exonic, and intronic sub-modes. We used 16 dispersion features of XY1 6-mer subsets to uncover the differences in the sequence composition among the six kinds of splicing modes. We found that the dispersion features can effectively distinguish the differences among the three modes and among the three sub-modes. Our study indicates that the dispersion features of 6-mer subsets can reveal the differences in base correlation in acceptor splicing sequences. In exploring base correlation, our method is important for studying functional sequences.

Identifying the sequence composition of different splicing modes is a challenge in current research. This study explored the dispersion distributions of 6-mer subsets in human acceptor splicing regions. Without differentiating acceptor splicing modes, obvious differences were observed across the upstream, core, and downstream regions of splicing sites for 16 dispersion distributions. These findings indicate that the dispersion value of each subset can effectively characterize the compositional properties of splicing sequences. When acceptor splicing sequences were classified into common, constitutive, and alternative modes, the differences in dispersion distributions for most of the XY1 6-mer subsets were significant among the three splicing modes. Furthermore, the alternative splicing mode was classified into normal, exonic, and intronic sub-modes, the differences in dispersion distributions for most of the XY1 6-mer subsets were also significant among the three splicing sub-modes. Our results indicate that dispersion values of XY1 6-mer subsets not only revealed the sequence composition patterns of acceptor splicing regions but also effectively identified the differences in base correlation among various acceptor splicing modes. Our research provides new insights into revealing and predicting different splicing modes.