Skip to content

Latest commit

 

History

History
74 lines (63 loc) · 2.98 KB

README.md

File metadata and controls

74 lines (63 loc) · 2.98 KB
Raw

Quick start guide

System setup

  1. Clone this repository
git clone https://github.com/abearab/QuantSeq-FWD-with-UMIs.git
  1. Create required conda environments as described here.

Preprocessing

Note: Before you run below commands, change '<#of-jobs>' to you desired number of threads to use for the analysis. Also, change '<parent-dir>' to the path which you want to write results and '<fastq-dir>' to the place which FASTQ files are located.

conda activate alignment 
cd ./src
bash fastqc.sh '<parent-dir>' '<fastq-dir>' fastQC '<#of-jobs>' 
bash process_fastq.sh '<parent-dir>' '<fastq-dir>' '<#of-jobs>'
bash alignment.sh '<parent-dir>' fastq-processed/trim '<#of-jobs>'
bash umi_dedup.sh '<parent-dir>' bam '<#of-jobs>'
bash htseq-count.sh '<parent-dir>' bam-processed counts
multiqc counts/ logs/ -n mutiqc-preprocessing

For a detailed explanation of all modules see this note.

Downstream analysis

Define a samplesheet.txt which contain a tab seprated table to describe your samples. [coming soon]

conda activate deseq2
Rscript DESeq.R '<parent-dir>' counts samplesheet.txt '<#of-jobs>'

Note: you can run each line with nohup in below format to run your command in the background:

nohup '<program>' '<options>' > '<log-file.txt>' &

Exploratory data analysis

Jupyter! Make sure to install Jupyter and nb_conda_kernels in the base environment or build seprate environment for that. For instance:

conda create -y -n nb-env
conda activate nb-env
conda install -y -c anaconda jupyter
conda install -y -c conda-forge nb_conda_kernels

Using nb_conda_kernels, you can have one Jupyter installed in your system and launch different python or R kernels for any created conda environments even in a single notebook. Therefore, you can run jupyther notebook command to launch Jupyther app and then, you can use your notebooks with kernels from different environments.

Note: You only need ipykernel, numpy and pandas in each environment in addition to your own packages.

conda install -n <env-name> -c anaconda ipykernel numpy pandas

How to use R (and python) in Jupyter?

If you want to use python and R packages together, you can use rpy2. After you have R installed through conda, you can install rpy2 through pip.

Instead, you can include R kernel into an envrinment with R packages. So, install irkernel.