Objective  To study codon usage bias of H9N2 avian influenza virus (AIV) complete genomes and its influence factors.

Method  The complete genomes of Chinese epidemic H9N2 AIV strains from 2010 to 2018 were selected. The characteristics of base composition, optimal codons, influence factors of codon usage bias and adaption to the codon usage patterns of the host were analyzed.

Result  AU content was higher than GC content in the whole genomes of H9N2 AIV. Most of the optimal codons ended with A or U, and the average of effective number of codons (ENC) was 52.86, suggesting that codon usage bias existed, but the bias was low. The codon usage bias was mainly affected by mutation pressure and natural selection. Natural selection (accounting for 61.79%−76.15%) played a greater role than mutation pressure (accounting for 23.85%−38.21%). In addition, the average codon adaptation ind ex of H9N2 AIV to Homo sapiens ranged from 0.739 to 0.741, suggesting that H9N2 AIV might have adapted to human codon usage patterns.

Conclusion  The study provides a theoretical basis for genetic evolution analysis of H9N2 AIV, codon optimization of existing vaccines and development of new vaccines (codon deoptimization vaccine).