Biobakey tools version for data reproducibility

I’ve used kneaddata version 0.12.0 (with the human genome hg37) for quality control on WGS data from a study, but noticed the authors originally used kneaddata version 0.7.2 with the same human genome. Should I redo the decontamination process using kneaddata v0.7.2, or is my current version sufficient?

Additionally, I’m looking for MetaPhlAn2/StrainPhlAn version 2.9.21 with the MetaPhlAn species-specific marker database version 2.94 (January 2019). Despite searching through the release history, I couldn’t find MetaPhlAn2/StrainPhlAn v2.9.21. Can you provide assistance?