The influenza A virus genome contains 8 gene segments encoding 10 commonly recognized proteins. Additional protein products have been identified, including PB1-F2 and PA-X. We report the in-silico identification of novel isoforms of PB1-F2 and PA-X in influenza virus genomes sequenced from avian samples. The isoform observed in PA-X includes a mutated stop codon that should extend the protein product by 8 amino acids. The isoform observed in PB1-F2 includes two nonsense mutations that should truncate the N-terminal region of the protein product and remove the entire mitochondrial targeting domain. Both isoforms were uncovered during automatic annotation of CEIRS sequence data. Nominally termed PA-X8 and PB1-F2-Cterm, both predicted isoforms were subsequently found in other annotated influenza genomes previously deposited in GenBank. Both isoforms were noticed due to discrepant annotations output by two annotation engines, indicating a benefit of incorporating multiple algorithms during gene annotation.

甲型流感病毒基因组包含 8 个基因片段，编码 10 种公认的蛋白质。已鉴定出其他蛋白质产物，包括 PB1-F2 和 PA-X。我们报告了in-silico在从禽类样本中测序的流感病毒基因组中鉴定新的 PB1-F2 和 PA-X 同种型。在 PA-X 中观察到的同工型包括一个突变的终止密码子，它应该将蛋白质产物延长 8 个氨基酸。在 PB1-F2 中观察到的亚型包括两个无义突变，它们应该截断蛋白质产物的 N 末端区域并去除整个线粒体靶向域。在 CEIRS 序列数据的自动注释过程中发现了两种同种型。名义上称为 PA-X8 和 PB1-F2-Cterm，这两种预测的同种型随后在之前存放在 GenBank 中的其他注释流感基因组中被发现。由于两个注释引擎输出的注释不一致，两种同种型都被注意到，这表明在基因注释期间结合多种算法的好处。