ANCHOR

Anchor is a 16S rRNA gene amplicon pipeline for microbial analysis of multiple environmental samples.

Getting Started

Installation with Docker

Docker

A pre-build docker image is available that includes all dependencies with the exception of usearch, which has a licence that precludes binary distribution. We recommend using our image and your own usearch binary to build a complete anchor container.

First, register your email addresss at https://drive5.com/usearch/download.html for USEARCH v9.2.64 and you will recieve an email with the download URL.

# Download the usearch binary
curl $URL_FROM_EMAIL > usearch9.2.64_i86linux32

# Make a tiny one-line Dockerfile
echo "FROM robsyme/anchor-onbuild" > Dockerfile.anchor-onbuild

# Build your Anchor Docker image.
docker build -t anchor -f Dockerfile.anchor-onbuild .

The robsyme/anchor-onbuild binary expects to find the file ./usearch9.2.64_i86linux32 when it is used as a base image to build a new container. It will grab the binary and fold it into your new docker image.

Installation Locally

Dependencies

Anchor requires the following tools:

• Mothur (used in assembling contigs. See: https://www.mothur.org/wiki/Installation)
• BLAST (see: https://www.ncbi.nlm.nih.gov/books/NBK279671)
• usearch9 (used for chimera detection. See: https://drive5.com/usearch/download.html)
• python 2.7 should be already installed on the machine (https://docs.python-guide.org/starting/install/linux)

... and the following Python libraries:

• numpy
• pandas
• matplotlib
• seaborn
• openpyxl
• biopython

Anchor

1. Download (or clone) ANCHOR_v1.0 from github (https://github.com/gonzalezem/ANCHOR/tree/master/ANCHOR_v1.0)
2. If not already within your system, create a link (or copy) of mothur main file into ANCHOR_v1.0/pipelineScripts/mothur. ANCHOR will look for a file called simply mothur within ANCHOR_v1.0/pipelineScripts/mothur/.
3. Create a link (or copy) of usearch9 main file (usearch9) into ANCHOR_v1.0/pipelineScripts/usearch9. ANCHOR will look for a file called simply usearch9 within ANCHOR_v1.0/pipelineScripts/usearch9/
4. Build (or link) database(s) BLAST index into ANCHOR_v1.0/db folder. Note that NCBI 16S microbial database index is included in ANCHOR download. The name should be: databasename_index (ex: 16SMicrobial_index, nt_index, rdp_index, silva_index) (see how to build an index: https://www.ncbi.nlm.nih.gov/books/NBK279688)

Running Anchor

Inputs

ANCHORS needs a few files and folders:

• A folder containing Illumina reads (ex. PEread1_R1.fastq.gz, PEread1_R2.fastq.gz, etc.)

• A design file containing at least 2 columns: Samples and Condition name. Example:

  Samples	myCondition
  PEread1	Condition1
  PEread2	Condition1
  PEread3	Condition1
  PEread4	Condition2
  PEread5	Condition2
  PEread6	Condition2

Before running ANCHOR, prepare some room for it. The script preparation_script.sh from within the ANCHOR folder will do this. This script will check for dependencies and required files. It needs 3 arguments to be able to run:

• argument 1: raw read location (full path)
• argument 2: folder from where ANCHOR will be run (full path)
• argument 3: design file (full path)

Example:

cd mycomputer/myfolder/ANCHOR_v1.0
bash preparation_script.sh myIlluminaFiles/my_raw_reads mycomputer/myExperiment mycomputer/myfolder/myconditions.txt

The script can be run multiple times until there is no more error message.

Run

If running the script preparation_script.sh didn't retrun an error, you're good to go. The last output lines from preparation_script.sh run will tell you what to do (basically customizing ANCHOR to your needs and running the main script).

bash bashMe.sh

Output

When anchor is done a folder Results_a_b_c_d will be created (a-d values depend on user's input from metadata/pipe.ini).

A few folders are produced:

• Summary (some summary files from ANCHOR run)
• STAMP (inut for STAMP software)
• Phyloseq (input for Phyloseq)
• MicrobiomeAnalyst (input for microbiomeanalyst.ca)
• metagenomeSeq (input for metagenomeSeq)
• Excel (OTU table in excel format) and files
• OTU and anchor sequences (fasta files)
• OTU and anchor tables (txt files)

Test Run

Go inside test_run folder and run the following command (it takes around 2 minutes to run):

bash run_test.sh

Figures From the Article

Figures here.