Isoseq_processing

Snakemake for aligning Isoseq reads and processing transcripts using Cupcake TOFU and SQANTI

For use in linux

Developed with snakemake (v3.13.3)

Create the conda environment

conda env create -f environment_isoseq.yaml

Test which version of R you are working with!

R (>= 3.4.0) R packages (for sqanti_qc2.py): ggplot2, scales, reshape, gridExtra, grid, dplyr

Install auxillary programs and create directory structure

Install Cupcake/ToFU (https://github.com/Magdoll/cDNA_Cupcake/wiki)

Basic steps:

conda activate isoseq_proc

cd ~/Tools/

git clone https://github.com/Magdoll/cDNA_Cupcake.git

cd ~/Tools/cDNA_Cupcake

python setup.py build

python setup.py install

Install TAMA

cd ~/Tools/

git clone https://github.com/mestill7/tama.git

Install SQANTI2

https://github.com/Magdoll/SQANTI2/

Basic steps:

cd ~/Tools

git clone https://github.com/Magdoll/SQANTI2.git

Install auxillary script (gtfToGenePred)

cd ~/Tools

wget http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/gtfToGenePred

Install Matchannot

https://github.com/TomSkelly/MatchAnnot.git

Basic steps:

cd ~/Tools

git clone https://github.com/TomSkelly/MatchAnnot.git

Once all prerequisites are installed:

Initialize the environment paths

conda activate isoseq_proc

export PYTHONPATH=$PYTHONPATH:~/Tools/cDNA_Cupcake/sequence/

export PYTHONPATH=$PYTHONPATH:~/Tools/cDNA_Cupcake/

module load R

export PATH=$PATH:~/Tools

The following commands should now work (error-free):

python ~/Tools/SQANTI2/sqanti_qc2.py --help

python ~/Tools/SQANTI2/sqanti_filter2.py --help

Create directory structure for snakemake process

For this example, let's say your directory is "~/isoseq"

Copy the following files into your "~/isoseq" directory,

such that your final directory structure is as follows:

.
├── config.yaml
├── cluster.json
├── run_snakemake_cluster.sh
├── run_snakemake.sh
└── Snakefile

Please modify the config.yaml as needed. Note that: Data files (e.g. polished isoseq transcripts and cluster reports) can be located within a different directory from the directory used to hold results. This functionality is intended to prevent unnecessary file duplications.

Running the Snakefile:

Make sure to initialize your environment paths as shown previously snakemake -nr --snakefile Snakefile ##shows what commands will be run

To use your LSF clusters:

nohup sh run_snakemake_cluster.sh &

To run Snakemake without an LSF cluster:

Type sh run_snakemake.sh followed by the maximum number of CPU cores to be used by snakemake. For example, type sh run_snakemake.sh 2 for 2 CPU cores. You can also type nohup sh run_snakemake.sh 2 & to run the pipeline in background.

Currently, the Snakemake file will combine the input fasta files and process the single file. Future implementations of the pipeline will allow individual fasta files to be processed independently.

Running the Snakemake process produces the results files from TAMA and runs SQANTI2 and matchannot.

Simplifying matchannot results

The matchannot program produces an extremely detailed report. However, users may only be interested in the associated MatchAnnot score for a given Isoseq isoform. In that case, the following awk script will extract the isoform name (1st column), associated known transcript (2nd, 3rd columns) and MatchAnnot score (4th column):

Enter directory holding your matchannot results first

awk 'BEGIN {OFS="\t"}; {if($1=="result:") print $2,$3,$4,$8}' Merged_samples_matchannot.txt > Merged_samples_matchannot_simplified.txt