准确的变异描述在遗传学领域至关重要。该领域面临日益复杂的变异类型，使得生成恰当的变异描述变得更加困难。我们提出了一种基于所有最小比对的图，它能够完整表示一个变异，并提供了三种互补的提取方法从该图中推导变异描述。实验结果显示，与dbSNP相比，我们的方法对于简单变异能生成相同的HGVS描述，而对于复杂变异则能生成更具意义的描述。

Accurate variant descriptions are of paramount importance in the field of genetics. The domain is confronted with increasingly complex variants, making it more challenging to generate proper variant descriptions. We present a graph based on all minimal alignments that is a complete representation of a variant and we provide three complementary extraction methods to derive variant descriptions from this graph. Our experiments show that our method in comparison with dbSNP results in identical HGVS descriptions for simple variants and more meaningful descriptions for complex variants.