Different taxonomic classifers will give different results, and the taxonomic classifier used to produce the source TAXID count table must be the same as the one used to produce the sink TAXID count table.
While there are many available taxonomic classifiers available to produce the source and sink TAXID table, the Sourcepredict author provide a simple pipeline to generate the source and sink TAXID table.
This pipeline is written using Nextflow, and handles the dependancies using conda. Briefly, this pipelines will firt trim and clip the sequencing files with AdapterRemoval before performing the taxonomic classification with Kraken2.
$ conda install -c bioconda nextflow $ nextflow pull maxibor/kraken-nf