Running the HMB_Taxa Pipeline

This set of scripts only works with indexed primers as specified in the sample sheet generator. It is also backwards compatible with the multiple versions of phasing previously used in primers.

You will need:

• USEARCH 8.0+ installed and in your PATH
• PEAR installed and in your PATH
• Copy these files to a working directory:
  • unzip_PEAR.py
  • HMB_Taxa.py
  • HMB_Database4.fa

Step 0 - Start a screen (optional):

Since these scripts can take a while to run, it is recommended to start a 'screen' by running:

screen

Step 1 - Unzip and Merge:

Copy the raw unzipped NGS data from BaseSpace or the MiSeq into a directory named 'zipped'. The scripts are compatible with both the MiSeq file structure and BaseSpace file structure. Call unzip_PEAR.py to unzip the files and merge the reads:

python unzip_PEAR.py zipped/

This will give you a live output of the unzipping/merging status. The outputs of this script are automatically saved into merge_report.log

The merged fasta files are outputted to a 'assembled' directory.

Step 2 - Assign Taxa and Generate Tally File:

Run:

python HMB_Taxa.py assembled/ HMB_Database4.fa

This script will first scan through each merged file and discard improper amplicons and also trim the phasing off proper amplicons. The filtered amplicons get put into a folder 'filtered' The script then runs USEARCH on each of the filtered files and tallies up the counts for each organism.

It outputs three files:

1. tally.txt

• This is the human-readable tally file.

2. tally.hmb

• This is a JSON formatted dictionary/array formatted form of the plate-counts mapping

3. report.txt

• This report contains statistics for each well and is a quick way to check for contamination.