Bioinformatics Analysis Of Human Alternative Splicing Based On Sequence Features

There is important significance in the study of codon bias and clustering analysis of alternative splicing. RSCU(relative synonymous codon usage) of constitutive exons and alternative exons were calculated in this study, it is found that the value of every codon is the same in two kinds of exons, except three stop codons(UAG, UGA, UAA). We also selected30constitutive exons and30alternative exons, and calculated RSCU values of these60sequences. Two kinds of exons were clustered based on RSCU and it is found that the two kinds of exons can not be distinguished. This result confirmed the codon biases are the same between the two kinds of exons. This result was the same as the one that we had directly compared RSCU values of the two kinds of exons. We calculated HI of the60sequences and two kinds of exons were clustered based on HI. It can’t be used to distinguish the two kinds of exons. Thus the two kinds of exons are not classified by RSCU value and HI value. It is suggested that regulated information of constitutive splicing and alternative splicing is seldom in the exons.The average values and standard deviations of D1and D2of constitutive exons and alternative exons are basically the same. Besides intron retention, the average values and standard deviations of D1and D2in the five kinds of alternative exons are basically the same. The values of D1and D2are significantly differences between intron retention and other four types of alternative exons. This result indicated that intron retention has characteristics of intronic sequences and exon sequences.In upstream of all kinds of exons,150bp-long sequences are cut out,and in downstream,150bp-long sequences are cut out, ID are calculated. All kinds of exons are cut out to two variants, ID are also calculated.It was found that the average ID values and the standard deviations of constitutive exons and alernative exons are basically the same in the two ways. Besides intron retention, the average ID values and the standard deviations of the other four kinds of exons are basically the same in the two ways. There are relatively large differences of the average ID values and the ID standard deviations between intron retention and other four types of exons.