The Amplicon Sequencing Analysis Pipeline (ASAP) is a highly customizable, automated way to examine amplicon sequencing data. The important details of the amplicon targets are described in a text-based input file written in JavaScript Object Notation (JSON). This data includes the target name, amplicon sequence (or sequences in the case of gene variant assays), any known SNPs or regions of interest (ROIs) within the target, and what the presence of this target or SNP signifies. The sequenced reads are processed by performing adapter, and optionally, quality trimming using Trimmomatic, and then aligned to the reference amplicon sequences extracted from the JSON file using one of several alignment packages (BWA-MEM, bowtie2, and NovoAlign are currently supported). The resulting BAM files are analyzed with a custom Python script using the pysam and scikit-bio libraries to aid in analysis. This script combines the alignment data in the BAM file with the assay data in the JSON file and interprets the results. The output is an XML file with complete details for each assay against each sample. These details include number of reads aligning to each target, any SNPs found above a user-defined threshold, and the nucleotide distribution at each of these SNP positions. For ROI assays, the output includes the sequence distribution at each of the regions of interest -- both the DNA sequences and translated into amino acid sequences. Also, each assay target is assigned a significance if it meets the requirements laid out in the JSON file (i.e. a particular SNP or amino acid change is present) To make this output easier for the user to interpret, an XSLT stylesheet is provided for transforming the XML output into a more readable set of web pages. Additionally, the use of XSLT stylesheets allows for multiple different views of the same data, from clinical summaries showing only the most important or relevant results to full researcher summaries containing all of the data.
- python 3.x
- pysam
- scikit-bio
- lxml
- samtools
- Trimmomatic (optional, but recommended)
- An aligner: either bowtie2 (default), Novoalign, or BWA-MEM
- PBS/Torque Cluster job manager
- Clone this git repository:
git clone [email protected]:TGenNorth/TB-ASAP.git - Use python3's easy-install script as root:
easy-install-3.x TB-ASAP - OR install to your user space:
easy-install-3.x --user TB-ASAP
*note: replace the 'x' with the appropriate version for your python3 installation
analyzeAmplicons -n <RUN_NAME> -j <INSTALL_DIR>/assay_data/Mtb6.json -r <DIRECTORY_OF_READ_FILES> -o <OUTPUT_DIR> --breadth 0
This will run TB-ASAP with the default paramaters, but turning off coverage breadth checking, since the reference sequences include flanking regions, submitting all of the jobs using PBS/torque and exiting. The final output file will be <OUTPUT_DIR>/<RUN_NAME>_analysis.xml. To see all of the additional options available, run analyzeAmplicons --help.
- Change into the
<OUTPUT_DIR>and run:formatOutput -s <INSTALL_DIR>/output_transforms/TB_RunResults_Mtb6.xsl -x <RUN_NAME>_analysis.xml
This will create HTML files: <RUN_NAME>.html and
<RUN_NAME>_details.html and a directory <RUN_NAME>/ of HTML
files for each sample. Open <RUN_NAME>.html in your web browser to
view the results.
Copyright ©️ The Translational Genomics Research Institute. See the included "LICENSE" document.
TGen North | 3051 W Shamrell Blvd Ste 106 | Flagstaff, AZ 86001-9435
Darrin Lemmer [email protected]