Creating a website has absolutely been a wild ride! I jumped between tutorials (and themes and hosts and generators) before finding the one I liked the best: Summer of Blogdown by Alison Hill.
The best way to improve prediction is to increase the amount of data you’re using to do the prediction. With warfarin, the anticoagulant I study, there are limited sources of explored variability in drug response.
Qiime My sequencing was done by Ion Torrent which made Qiime a bit difficult to use because of lacking documentation. I followed, mostly, this post by Carli Jones: https://forum.qiime2.org/t/possible-analysis-pipeline-for-ion-torrent-16s-metagenomics-kit-data-in-qiime2/13476.