Cyclon trxtools example

Supplemental code for XXXXXX et al. 20XX. This repository provides step-by-step guidance to reproduce data in the MS and serves as a manual for using the trxtools package. The workflow is divided into two stages (cluster and desktop) according to computational demand, however it can run entirely on as single machine. All steps require git and conda installed.

Cluster processing

Getting started

Clone this repository

git clone git@github.com:TurowskiLab/example-trxtools-Cyclon.git

Prepare STAR index

conda env create -f envs/processing.yml
conda activate processing
STAR --runThreadN 30 --runMode genomeGenerate --genomeDir hg41_STAR_index/ --genomeFastaFiles GRCh38.primary_assembly.genome.cleaned.fa --sjdbGTFfile gencode.v41.annotation.gtf --limitGenomeGenerateRAM 33524399488

NOTE: Adjust --runThreadN (number of CPU threads to use) and --limitGenomeGenerateRAM (available RAM) to your systems capabilities.

NOTE: The STAR index can be saved to any location (--genomeDir).

IMPORTANT: You need to specify the path to your STAR index in the Snakefile. Set STAR_INDEX to an absolute path to your index, e.g. STAR_INDEX = "/home/user/seq_references/hg41/hg41_STAR_index/"

Create and activate conda environment (you can use mamba instead)

conda env create -f envs/snakemake.yml
conda activate snakemake

Run Snakemake file to process raw files

Try a dry run first:

snakemake -c64 --use-conda -n

NOTE: -c determines number of CPUs to use.

If no errors are reported you can start the run proper:

snakemake -c64 --use-conda

OR if using slurm:

snakemake  -c64 --use-conda --slurm -j12

NOTE: The first run of the Snakemake file will initialize new conda environments. This may take several minutes.

After the run finishes continue to the analysis steps below. If you're performing the analysis stage on a different machine (e.g. a desktop after running the pipeline on a cluster), copy the whole repository folder there, including the output files produced.

Analysis using Jupyter notebooks

Create and activate jupyter environment

conda env create -f envs/jupyter-trxtools.yml -n jupyter-trxtools
conda activate jupyter-trxtools

Open Jupyter Lab and run notebooks

jupyter lab .

Afterwards open the subsequent notebooks in Jupyter and run them to perform the analysis.

Authors

Jan Mikołajczyk, Tomasz W. Turowski