The classify.seqs command allows the user to use several different methods to assign their sequences tothe taxonomy outline of their choice. Current methods include using a k-nearestneighbor consensus and Wang approach.  

Classify 用两种方法来计算分类：knn 和wang. 两者的区别是：knn快，不计算supportivevalue. Wang method 计算。

这里需要对比的序列和分类的文件：

http://www.mothur.org/wiki/Silva_reference_files  这个是常用的silva，还有greengene

mothur >classify.seqs(fasta=abrecovery.fasta, template=nogap.bacteria.fasta,taxonomy=silva.bacteria.silva.tax)

wang method的defaultmethod,，根据query sequences里面kmer比对taxonomy，default ksize是8，兼顾快速和准确度。同时迭代次数也可以设定，default为100 ，cutoff参数去除supportive value 小于设定的序列。