##Requirements
- Linux (Arch Linux and Ubuntu Server 13.10 used). OS X may run the needed software, but is untested.
- A large amount of memory. Approximately 137GB RAM is used during the get.oturep() step, and ~47GB is used during the cluster() step. Amazon EC2 instance type cr1.8xlarge was used for memory intensive steps, which provides 244GB of memory.
- Trimmomatic (version 0.32 used)
- Mothur (version 1.32.1 used)
- Python 2 and 3 (versions 2.7.6 and 3.3.3 used)
- BioPython (version 1.63 used)
- R (version 3.0.2 used)
- R libraries ggplot2, plyr, gridExtra
- Krona
##Obtaining raw reads Raw reads are deposited in Zenodo (http://dx.doi.org/10.5281/zenodo.11120).
##Cleaning raw reads Prior to running mothur, raw fastq sequences need to be cleaned with trimmomatic. A convenience script is provided at scripts/print_trimm_cmd.py if changes to path names need to be made. By default, the scripts assume the raw reads are in fastqs_raw/
$bash scripts/trimm_raw_reads.sh
##Running mothur
$mothur protocol.batch
Annotations are provided in the mothur batch file, explaining steps. The mothur protocol also runs accessory scripts, with additional requirements as seen above.
##A note on sample naming Sample and sequences have abbreviated names during data analysis.
ICfXX_YY: INT-Coffee, Sample XX, Sequence YY (only for seqences)
ICn: INT-Control
TCf: TRN-Coffee
TCn: TRN-Control
OCf: ORG-Coffee
OCn: ORG-Control
###Contact info primary email: adam.caldwell@gmail.com