Ananke_PeerJ

Scripts useful for reproducing the results in the Ananke manuscript found in PeerJ PrePrints (currently under revision for PeerJ)

Two data-sets were processed with Ananke in this publication, a freshwater lake data set and a stool dataset. To reproduce the analyses in the publication, pull the scripts with git clone https://github.com/mwhall/Ananke_PeerJ.git.

Required Software

• ananke (v0.3)
• ananke-ui
• R (v3.4.0)
• DADA2 (v1.4)
• lubridate (v1.6.0)
• awk
• curl
• wget

Freshwater Lake

Data is from the North Temperate Lakes Long-term Ecological Research Network (https://lter.limnology.wisc.edu/). Sequences are at EBI project accession PRJEB14911.

Scripts are found in the McMahon_Mendota directory. There are two analyses possible: one with DADA2 denoising, and one without.

With DADA2 denoising, there are 4 scripts that must be run in the presented order:

• fetchData.sh
  • BASH script that communicates with EBI to download the .fastq.gz files directly, as well as all associated sample metadata.
• runDADA2.R
  • R script that runs DADA2 v1.4 to denoise the sequences, remove PhiX, and filter out bimeras.
• fixDates.R
  • R script that uses lubridate to convert the dates in the metadata to integer offsets from the earliest date (a requirement for Ananke v0.3). It also removes low sequence-depth samples from the metadata file, which excludes them from the Ananke analysis.
• TaxAss.sh
  • Runs the TaxAss pipeline with FreshTrain to classify the sequences from DADA2. See https://github.com/McMahonLab/TaxAss for more information. Pre-computed taxonomic classifications are given in taxonomy.txt.
• Ananke.sh
  • BASH script that runs the Ananke commands to import DADA2 results, cluster, and then import the taxonomic classifications and the sequence-identity clusters.

Without DADA2 denoising, there are 3 scripts that must be run in the presented order:

• fetchData.sh
  • BASH script that communicates with EBI to download the .fastq.gz files directly, as well as all associated sample metadata.
• fixDates.R
  • R script that runs DADA2 v1.4 to denoise the sequences, remove PhiX, and filter out bimeras.
• TaxAss.sh
  • NOTE: Some code comments need to be swapped for the no-DADA2 full sequence run. This script runs the TaxAss pipeline with FreshTrain to classify the raw sequences. See https://github.com/McMahonLab/TaxAss for more information. Pre-computed taxonomic classifications are given in taxonomy_full.txt.
• Ananke_all_sequences.sh
  • BASH script that runs the Ananke commands to tabulate sequences, cluster, and then import the taxonomic classifications and the sequence-identity clusters.

Stool

Data is from David et al., 2014 (https://genomebiology.biomedcentral.com/articles/10.1186/gb-2014-15-7-r89). Sequences are at EBI project accession PRJEB6518.

Scripts are found in the Alm_Stool directory. There are 3 scripts that must be run in the presented order:

• fetchData.sh
  • BASH script that communicates with EBI to download the .fastq.gz files directly, as well as all associated sample metadata.
• runDADA2.R
  • R script that runs DADA2 v1.4 to denoise the sequences, remove PhiX, and filter out bimeras.
• Ananke.sh
  • BASH script that runs the Ananke commands to import DADA2 results, cluster, and then import the taxonomic classifications and the sequence-identity clusters.