RDP classifier - training set: how do I create?

[bio_informatics]

Hi there,

I've nasal and genital microbiome: 16S rRNA, Illumina 300BP, paired end data. For various reasons green-genes, and silva database don't solve our project's purpose, and thus we don't use them.

I'd like to create in-house training set. I've list of species which we're interested in.

How do I create RDP's training set for my species? I looked at files:

genus_wordConditionalProbList.txt
rRNAClassifier.properties
bergeyTrainingTree.xml
logWordPrior.txt

these files make almost no sense to me, and am clueless on how to add new sequence, classification/taxonomy to it.

Any guidance would be great.