PREMIX

The primary motivation for this repository to support the reproducibility of the analyses from the PREMIX study.

At each step in analysis, we post the tools, versions, and annotated code used with (in most cases) the expected input and output for metagenomic data from human samples with or without paired data from cultured isolates. We use the genomic and metagenomic analyses from the paper Randomized trial: Fecal microbiota transplant promotes reduction of antimicrobial resistance by strain replacement as a practical example.

Data for the analyses in the PREMIX paper that will be used for this repository are available through the NCBI Bioproject (accession PRJNA728680). Other data and some intermediate analytic output are also available through the PREMIX Emory DataVerse site.

---

Overview

• Background
• Quality Control
  • (human) stool metagenomes
  • isolate genomes
• Metagenomic Analyses
  • Assembly (metaSPAdes)
  • Short-read taxonomic classification (kraken2/bracken)
  • Metagenome-assembled genome binning & classification (maxbin2 / metabat2 / DASTool / gtdbtk)
• Genomic Analyses
  • Assembly (SPAdes)
  • Gene prediction (Prodigal)
  • Antimicrobial resistance genes analyses (AMRFinder)
  • Annotation (prokka | MicrobeAnnotator)
  • Clustered gene analyses (Roary)
  • Average nucleotide identity (FastANI)
• Integrated Genomic and Metagenomic Analyses
  • Assemble contigs (see steps in metagenomic analyses)
  • Predict proteins (Prodigal)
  • Annotate genes (with AMRFinder Plus as motivating example)
  • Map metagenome reads vs reference (Bowtie2)
  • Filter reads and estimate coverage and diversity statistics (inStrain)
• Visualization