MVIPER: Bulk_RNA-Seq_Workflow_Snakemake

STAR, cufflinks, cluster, DESeq2, GO analysis.

---

This workflow was designed to process and visualize bulk RNA-seq data. The output files including:

• Mapping reads with STAR;
• Counting reads with STAR and cufflinks;
• Samples quality control with PCA plot and samples-samples clustering heatmap;
• DEG with DESeq2 and limma;
• Functional enrichment analysis with GO and GSEA.

---

Table of Contents

1. MVIPER
2. Working directory structure
3. How to run the MVIPER
4. Running VIPER
5. Outputs of MVIPER

MVIPER

MVIPER is a bulk RNA-seq analysis pipeline built using snakemake. MVIPER is modified VIPER. Modifications are as the follows:

• add scripts to generate config.yaml and metashee.csv automatically
• add script to generate gene annotation file which is used in the ref.yaml for new species
• modified file_format.snakefile to transfer bam file to BigWig file format using deeptools
• modified preprocess.snakefile to set multiple thresholds for top variant genes lists
• modified DE.snakefile to set multiple thresholds for DEG selection
• modified serveral R scripts and python scripts to get more beautified plots

Cornwell M, Vangala M, Taing L, Herbert Z, Köster J, Li B, Sun H, Li T, Zhang J, Qiu X, Pun M, Jeselsohn R, Brown M, Liu XS, Long HW. VIPER: Visualization Pipeline for RNA-seq, a Snakemake workflow for efficient and complete RNA-seq analysis. BMC Bioinformatics. 2018 Apr 12; 19(1):135. PMID: 29649993.

Working directory structure

Download the MVIPER by the following command: git clone https://github.com/YutingPKU/Bulk_RNA-Seq_Workflow_Snakemake.git

Put the MVIPER and your input data in a directory (like PROJECT).

PROJECT/ - the root directory modules/ - the scripts and snakefile for MVIPER static/ - the reference metadata for MVIPER data/ - the input data directory config.yaml - pipeline configure file ref.yaml - reference metadata configure file metasheet.csv - metadata for input data samples Note: the data directory is the pathway of your input data, you can generate a soft link to your fastq files. Example is listed in the following ln -s fastq_directory data

How to run the MVIPER

1. setting the ref.yaml

• star_index: the genome index directory for STAR alignment
• gtf_file: gene annotation file by gtf file format
• gene_annotation: csv file containing gene symbol, gene description, ENSEMBL id, EntreZ id, GO id and GO term. For human, mouse and macaque, you can use the bz2 file from the static/ directory. For other species, you can generate the gene annotation file by the following command: Rscript step0_generateGeneAnnoFiles_snakemake.Rstep0_generateGeneAnnoFiles_snakemake.R Note: change the biomart dataset according to your species

2. setting the config.yaml

• set the configure to control the pipeline running and scripts parameter
• adding the sample names and pathways(fastq file directory) of the input data to the samples key. Examples are listed in the following

 samples:
    10068A-CPi-RNA-lib:
      - data/10068A-CPi-RNA-lib/10068A-CPi-RNA-lib_R1.fq.gz
      - data/10068A-CPi-RNA-lib/10068A-CPi-RNA-lib_R2.fq.gz

You can add the samples automatically by the following command bash addDataInfo_configymal.sh

3. setting the metasheet.csv

• SampleName: sample names of the input data, must be the same as config.yaml
• You can add any annotation information for the samples by adding in the columns
• Comparison group information are given by columns which start with "comp_", 1 means controls, 2 means treatments. You can set multiple comparison groups. Examples are listed in the following

SampleName	Regions	Replicates	comp_CPivsOther	comp_CPovsOther
10068A-CPi-RNA-lib	CPi	10068A	2	1
10068A-CPo-RNA-lib	CPo	10068A	1	2

4. test and run the workflow

• validate the pipeline by snakemake --np -s viper.snakefile
• run the pipeline by snakemek -s viper.snakefile --cores 20 -j 10
• run the pipeline in cluster by snakemake -j 10 -pr -c "pkubatch -p cn-short -N 1 -c 20 --qos=lch3000cns -A lch3000_g1 -J {rule}.{wildcards} -o logs/cluster/{rule}/{rule}.{wildcards}_%j.out -e logs/cluster/{rule}/{rule}.{wildcards}_%j.err " -s viper.snakefile -k 2 Note: setting the account information according to your cluster account

##Outputs of MVIPER benchmarks files and log files are in the bednchmarks and logs directories. Analysis results and Visualization resutls are in the analysis directory. The Contents of analyais directory are listed in the following

bam2bw/ - BigWig files of the aligned bam files per sample
cufflinks/ - genes and isoforms fpkm matrix per sample
STAR/ - STAR alignment results per sample
summary_reports/
  cufflinks/ - genes fpkm matrix for all samples; top variant genes fpkm matrix for all samples
  STAR/ - alignment reports for all samples
  plots/ - clustering heatmap based on samples-samples correlation matrix; PCA plots based on top variant genes fpkm matrix
  diffexp/ - DEG lists detected by DESeq2 and limma; Vocalno plots for DEG
  GO/ - GO, KEGG Pathway and GSEA analyses and Visualization for DEG

Contact

If you have any questions about MVIPER, please feel free to contact lyt17@pku.edu.cn