TC_hunter

TC-hunter identifies transgenic insertion sites within host genome

TC-hunter searches for transgenic insertion sites in a host genome and returns figures and a report to support these findings.

There's two programs; TC_hunter and TC_hunter_BWA.

• 🟩 TC_hunter_BWA.nf

TC_hunter_BWA accepts raw pair end fastq files (from one or several samples) as input and performes BWA MEM alignment before searching for trasgenic insertion site.

• 🟨 TC_hunter.nf

Accepts one or several aligned BAM files (mapped to both host and transgenic sequence) as input. TC-hunter then identifies anchors and chimeric reads that maps to both host and transgenig sequence.

Install TC-hunter

Clone the repository from Github and put it in your path (or add the direct path to config file)

$ git clone https://github.com/vborjesson/TC_hunter.git
$ export PATH="/home/yourPath/TC_hunter":$PATH

Software Dependencies

In order to run TC_hunter you need to have some programs installed. Here's three options on how you can do it:

1. Install required programs and tools using conda yml-file (prefered). Has been tested on Anconda3, Anaconda2 and Miniconda2

$ conda env create --file TC_hunter/Scripts/TC_hunter.yml
$ source activate TC_hunter_v1.0

2. Create your own conda environment

$ conda create -n TC_hunter R=3.5
$ source activate TC_hunter
$ conda install -c bioconda samtools=1.10
$ conda install -c bioconda nextflow=19.01.0
(only if runing TC_hunter_BWA) $ conda install -c bioconda bwa
$ conda install -c anaconda pandas
$ conda install -c conda-forge r-circlize
$ conda install -c r r-dplyr
$ conda install -c r r-data.table

3. Download manually

softwares

R 3.5 or higher
python 2.7
samtools 1.10 (works on other versions as well)
nextflow 19.01.0
bwa 0.7

R packages

circlize
dplyr
data.table

Run TC_hunter with test data (takes approximately 1 minute to run)

Download data

mkdir test_run 
cd test_run
pip install gdown # If you don't already have it installed
gdown https://drive.google.com/drive/folders/1Y-iCNo71OVmf3QqJeFrukxUlQGDojSKx?usp=sharing
cp ../TC_hunter/Test_data/* .

Then run TC_hunter:

nextflow ../TC_hunter/TC_hunter.nf -c testrun.config --workingDir <realpath_to_test_run_dir> --tc_hunter_path <realpath_to_tchunter>

You should see TC_hunter running each process one after each other

1. samtools_index
2. create_links_sup
3. create_links_soft
4. create_karyotype
5. create_histogram
6. create_plots
7. create_html

When it's done check that you have an output_summary.html file.

Create construct.txt file (required)

In order to generate figures with construct information, you need to add this informtaion. Create a txt-file with gene info per line, separated by space. The info should be; 1) name, 2) start position and 3) end position.

e.g.

Amp 1 500   
lyz 1000 1200       
Gene3 2000 5000		      
Gene4 7000 7700	     		

Make Configuration file

Create a configuration file from template.

$ cp TC_hunter/template/TC_hunter.config /path/to/WorkingDir 

Add required information to config file

TC_hunter.nf

Argument	Usage	Description
WorkingDir	<Path/WorkingDir>	Path to your working directory (this is where the output html and figures will be)
TC_hunter_path	<Path/TC_hunter>	Path to TC_hunter, only TC_hunter if it's in your $PATH
Construct_file	<Path/construct.txt>	Path to your construct.txt file (See Create construct.txt file above)
Construct_length		The length of your construct in numbers
Construct_name		The name of the construct, most match the name in the reference file, no space
bam	<Bam_directory>	The path to the directory where you have your bam file or (if several sampes) bam files.
Reference	<Jointref.fa>	Path to the merged reference file including both host and construct genome. cat host_ref construct_ref > Jointref.fa

e. g. example.config

TC_hunter_BWA.nf

Argument	Usage	Description
WorkingDir	<Path/WorkingDir>	Path to your working directory (this is where the output html and figures will be)
TC_hunter_path	<Path/TC_hunter>	Path to TC_hunter, only TC_hunter if it's in your $PATH
Construct_file	<Path/construct.txt>	Path to your construct.txt file (See Create construct.txt file above)
Construct_length		Length in numbers of your construct that will be plotted
Construct_name		Name of the construct, most match the name in your reference file
sample	<sample_directory>	Path to directory where you have the fastq-files (needs to have the name 'R1' and 'R2')
folder	<sample_directory>	Path to directory containing one directory for each sample. The name of the samples will be the same as the directory names
host_ref	<host_ref.fa>	Path to host reference file
construct_ref	<construct_ref.fa>	Path to construct reference file

e. g. example.config

Run TC_hunter.nf

Before running, make sure you have a config file with all required information (see "Make Configuration file").

$ nextflow TC_hunter.nf -c <file.config> [-with-report <report name>]

Run TC_hunter_BWA.nf

Before running, make sure you have a config file with all required information (see "Make Configuration file").

$ nextflow TC_hunter_BWA.nf -c <file.config> [-with-report <report name>]

Run IGV separately

In order to get the IGV figures you need to have GUI available. If not; you can run IGV separately when TC-hunter is finished. Run one .bat file for each sample.

$ igv.sh -b <sample_name.bat>

Understand your output

TC-hunter finds insertion sites based on chimeric and discordant read pair.

TC_hunter reports each possible insertion site in an html file called output_summary.html. The file contains 5 columns; 1) Ranking - best hit based on score is ranked first, second best second etc, 2) Score - Based on the number of chimeric and discordant read pairs supporting this insertion site, 3) Breakpoint host - Where in the host is this insertionsite located, 4) Breakpoint construct - Where in the construct is this insertion site located, 5) figures - three figures I) circular plot (see below), II) igv, III) igv more zoomed in.

• output_summary.html

For every predicted insertion site a circular figure is created. Red links, "lines" represent every discordant read pair supporting this event. Black links represent chimeric reads supporting this event. !

Supporting material for TC_hunter paper

Supporting data