Skip to content

Latest commit

 

History

History
194 lines (157 loc) · 7.79 KB

README.md

File metadata and controls

194 lines (157 loc) · 7.79 KB
Raw

CardioPipeLine

This is a simple utility to analyse RNA-seq data with a selection of popular tools. It is written in snakemake that allows the user to scale it to any system, including clusters and cloud computing architectures. Currently, it supports analyses with the following tools;

In short, given a public run ID from the SRA archives (SRRXXXXXXX), the pipeline can download the raw data from the NCBI server, convert the files to fastq format, filter the reads, generate indices, map to genome/transcript, and quantify the mapped reads. In addition, it can also carry out differential expression analysis.

However, the user can provide their own FASTA or SRA files for the analysis. In order to modify the default options, refer to the appropriate rules in the respective rules/filename.smk files. Please read wiki for more details.

cardio

Dependencies

  • python3+ (tested on Python3.7 and 3.8)
  • snakemake (workflow management)
  • pandas
  • R 4+ (for DESeq2 analysis; tested on R 4.0.2)
  • tximport, DESeq2, AnnotationDbi, Rsubread (for R related analyses)

All the above-mentioned dependencies have been successfully tested on Ubuntu 18.04.5 , Gentoo 5.4.48, Fedora 32.

Installation

For installation purposes, it is a good idea to make a new python virtual environment and install snakemake on it, and then download all the tool binaries in your desired location. Following this, retrieve the repository using git;

git clone https://github.com/rohitsuratekar/CardioPipeLine.git

Alternatively, download the repository in the desired location.

The only other modification that needs to be done is to edit the config/config.yaml and config/samples.csv files to specify the location of files on your system, and then run following in the repository folder

snakemake --core all

The above mentioned command will essentially carry out all the steps for you, from downloading the file to quantification, while you sit back and relax :)

Tasks

This pipeline carries out tasks. However, you can always select which tasks you want :)

Task No Details Tool Used Input # Output #, *
0 All tasks -- -- --
1 Download .sra file prefetch samples.csv srr.sra
2 Convert to fastq fasterq-dump srr.sra srr.fastq
3 rRNA filtering sortmerna srr.fastq srr.filtered.fastq
4 STAR indexing star genome.fa, anno.gtf star-idx
5 STAR mapping star star-idx, srr.filtered.fastq / srr.fastq srr.bam
6 Quantification stringtie srr.bam, anno.gtf st.tsv
7 Salmon indexing salmon anno.gtf salmon-idx
8 Salmon mapping salmon salmon-idx, srr.filtered.fastq / srr.fastq quants.sf
9 Kallisto indexing kallisto anno.gtf kallisto-idx
10 Kallisto mapping kallisto kallisto-idx, srr.filtered.fastq / srr.fastq abundance.tsv
11 Count Matrix StringTie prepDE.py st.tsv stringtie.counts
12 Count Matrix Star Rsubread ssr.bam star.counts
13 Count Matrix Salmon tximport quants.sf salmon.counts
14 Count Matrix Kallisto tximport abundance.tsv kallisto.counts
15 Differential Analysis DESeq2 *.counts con1_vs_con2.csv
16 Quality Control Report fastqc -- srr.html
17 Quality Control Report multiqc -- quality_report.html
18 de-novo Assembly trinity srr.fastq trinity.fasta
19 de-novo Alignment bowtie2 trinity.fasta stat.txt

* There might be many other related outputs. # Names are for representative purpose. Actual names will be different

Selection of tasks

You have full control of which tasks to run. Your can run desired tasks by editing tasks parameter in the config.yaml file. If you want to perform specific task, you need to provide its input file in appropriate location.

But do not worry, thanks to snakemake, if you did not provide appropriate input files, it will search far task which outputs these files and will automatically run for you.

By default tasks: 0 is set. This will run all available tasks.

Setting up the workflow

Editing config.yaml and samples.csv is probably the most important aspect of this pipeline. These are files which will change for each user according to their system and work.

config.yaml

If you are fine with default options of tools used in this workflow, you can just edit this file and run the pipeline with snakemake. Just read the comment in config/config.yaml file and you will know what to do :) Comments are self explanatory.

Config file has parameter called base which provides base folder for all the analysis. Every output file crated with this pipeline can be found in this base folder. Overall output file structure is shown below. More detailed file structure can be found here.

base
├── sra
├── fastq
├── filtered
├── bams
├── index
├── mappings
├── deseq2
└── logs

samples.csv

This file will contain names of all the samples which you want to analyse with this pipeline. You should provide SRA Runs IDs (usually in SRRXXXXXXX format ). You can find these IDs for publically available datasets from GEO Dataset website. If you have your own fastq files, you can check naming convention and rename those files and keep in appropriate folder. and start the pipeline.

This .csv (comma delimited) file SHOULD have header row and SHOULD contains at least following headers.

  • run : Run ID
  • is_paired : true (if you are handling paired end sample), false (if it is single end sample)

Example file content

run,is_paired
SRR0000001,true
SRR0000002,false

If you are performing any DESeq2 analysis, third column specifying conditions is mandatory

run,is_paired,condition
SRR0000001,true,wt
SRR0000002,false,mt

Here third column condition will be used in DESeq2 analysis. You should update appropriate parameters in config.yaml file. For example, Let us assume following sample file

run,is_paired,time
SRR0000001,true,48
SRR0000002,false,24

In above file, as third column name is time and 24 is your reference condition then you should update following in the config.yaml

deseq2:
    design_column: "time"
    reference: "24"

In addition, you should also update design parameter from config.yaml file. For example, in above situation, it can be design: "~ time"