RNASeq-Pipeline is a bioinformatics pipeline that can be used for analyzing RNA Sequencing data. RNASeq-Pipeline steps;
- Quality Control with FastQC and reporting with MultiQC
- Trimming with Trim-Galore
- Mapping with STAR
- Quantification with Salmon
- Differential Gene Expression Analysis with DESeq2
- Enrichment Analysis with pathfindR
All required packages can be installed using setup.sh script. setup.sh script creates a conda environment called rnaseq-pipeline and downloads all required packages in it.
bash setup.sh
For using the pipeline, conda environment should be activated.
conda activate rnaseq-pipeline
All arguments can be seen by using --help command.
python rnaseq-pipeline.py --help
usage: rnaseq-pipeline.py [-h] -s [...] -c [...] [-r] [-m [...]] [-t ] [-o ] [-f ] -d -R -G -x [-T ] [-q ] [-l ] [-i ] [-g ] [-M ] [-n ] [-B ] [-y ] [-p ] [-N ] [-F ] [-L ] [-P ] [-I ] [-D ]
RNASeq-Pipeline for Quantification and Enrichment Analysis
optional arguments:
-h, --help show this help message and exit
-s [ ...], --samples [ ...] Sample names
-c [ ...], --controls [ ...] Control names
-r , --replicate Technical replicate condition = YES or NO (default)
-m [ ...], --replicate_samples [ ...]
if --replicate == YES; Comma seperated matched sample names; sample1_sample2, sample3_sample4
-t [], --type [] Sample type; paired (default), single
-o [], --output [] Final output; pathway_enrichment (default), trimming, mapping, quantification, diff_exp
-f [], --is_fastqc [] Run FastQC or not; YES (default), NO
-d , --data_dir Data folder direction
-R , --reference_genome Reference genome path
-G , --gtf_file GTF file path
-x , --star_index Folder path to create genome index (if already exist, automatically use the indexed genome in the path)
-T [], --threads [] Number of thread (default = 16)
-q [], --quality [] Trim galore quality option (default = 20)
-l [], --length [] Trim galore length option (default = 35)
-i [], --genomeSAsparseD [] genomeSAsparseD argument for STAR (default = 2)
-g [], --runMode [] runMode argument for STAR (default = genomeGenerate)
-M [], --limitGenomeGenerateRAM [] Available maximum RAM size in bytes (default = 164282414959)
-n [], --genomeSAindexNbases [] genomeSAindexNbases argument for STAR (default = 14)
-B [], --genomeChrBinNbits [] genomeChrBinNbits argument for STAR (default = 18)
-y [], --outSAMtype [] Output SAM file type (default = BAM Unsorted)
-p [], --twopassMode [] twopassMode argument for STAR (default = Basic)
-N [], --twopass1readsN [] twopass1readsN argument for STAR (default = -1)
-F [], --readFilesCommand [] readFilesCommand argument for STAR (default = zcat)
-L [], --library_type [] Library type for Salmon (default = A)
-P [], --p_val_threshold [] p-value threshold for pathfindR (default = 0.05)
-I [], --iterations [] iteration number for pathfindR (default = 25)
-D [], --create_DAG [] Create Directed Acyclic Graph (DAG) of the workflow and exit (default = NO)
RNASeq-Pipeline can only works with .fastq.gz files. Input names should be;
- For pair-end samples:
{sample_name}_{lane = 1 or 2}.fastq.gz - For single-end samples:
{sample_name}.fastq.gz
If reference genome indexed before, RNASeq-Pipeline automatically uses the indexed genome path in the --star_index argument. If not, index genome will be created into the --star_index path.
python rnaseq-pipeline.py --samples sample1 sample2 sample3 \
--controls control1 control2 control3 \
--data_dir /path/to/data/ \
--reference_genome /path/to/reference_genome \
--gtf_file /path/to/gtf_file \
--star_index /path/to/target/folder
Output of the pipeline can be specified with using --output argument.
If you want to do differential expression analysis as a last step, you can set --output argument to diff_exp and pipeline automatically stops after that step.
python rnaseq-pipeline.py --samples sample1 sample2 sample3 \
--controls control1 control2 control3 \
--data_dir /path/to/data/ \
--reference_genome /path/to/reference_genome \
--gtf_file /path/to/gtf_file \
--star_index /path/to/target/folder \
--output diff_exp
Directed Acyclic Graph (DAG) of the analysis can be created by using --create_DAG argument.
python rnaseq-pipeline.py --samples sample1 sample2 \
--controls control1 control2 \
--data_dir /path/to/data/ \
--reference_genome /path/to/reference_genome \
--gtf_file /path/to/gtf_file \
--star_index /path/to/target/folder \
--create_DAG YES
If there are technical replicates in the samples, --replicate argument should be set as YES and technical replicates should be written with using --replicate_samples argument. RNASeq-Pipeline automatically combines technical replicates when running DESeq2.
python rnaseq-pipeline.py --samples sample1 sample2 sample3 sample4 \
--controls control1 control2 control3 control4 \
--replicate YES \
--replicate_samples sample1_sample2 sample3_sample4 control1_control2 control3_control4 \
--data_dir /path/to/data/ \
--reference_genome /path/to/reference_genome \
--gtf_file /path/to/gtf_file \
--star_index /path/to/target/folder
All the outputs will be written in seperate folders in the data folder for each step.
├── sample1_1.fastq.gz
├── sample1_2.fastq.gz
├── sample2_1.fastq.gz
├── sample2_2.fastq.gz
├── control1_1.fastq.gz
├── control1_2.fastq.gz
├── control2_1.fastq.gz
├── control2_2.fastq.gz
└── results
├── quality_control
│ └── multiqc_report.html
├── transcript_file_salmon
│ └── all_transcript.fa
├── trimming
│ ├── sample1
│ │ └── trimmed files
│ ├── sample2
│ │ └── trimmed files
│ ├── control1
│ │ └── trimmed files
│ ├── control2
│ │ └── trimmed files
├── mapping
│ ├── sample1
│ │ └── bam files
│ ├── sample2
│ │ └── bam files
│ ├── control1
│ │ └── bam files
│ ├── control2
│ │ └── bam files
├── quantification
│ ├── sample1
│ │ └── quantification result
│ ├── sample2
│ │ └── quantification result
│ ├── control1
│ │ └── quantification result
│ ├── control2
│ │ └── quantification result
├── differential_expression
│ ├── gene_name_converter.csv
│ ├── sample1_control1
│ │ ├── count_table.csv
│ │ └── deseq_output.csv
└── enrichment_analysis
├── enrichment_result.csv
└── active_snw_search
└── iterations