A comprehensive pipeline for small RNAseq data analysis.
This pipeline has been used in the following publication:
-
Ye, X., Yang, Y., ..., Sun, Y.H., ..., 2022. Genomic signatures associated with maintenance of genome stability and venom turnover in two parasitoid wasps. Nat Commun, 13, 6417
-
Liu, Z., Zhang, X., ..., Sun, Y.H.$ and Huo, J.$ 2023. Long- and short-read RNA sequencing from five reproductive organs of boar. Sci Data, 10, 678 ($co-corresponding authors)
Some folder/files are shared with other Pipe* pipelines (i.e. PipeRNAseq and PipeRiboseq), such as the PipelineHomeDir folder, and some scripts in the bin folder.
This pipeline is designed to run on Linux servers, and requires the following softwares. They need to be installed and added to the $PATH before using the pipeline.
R
Python3
STAR
bowtie
bedtools
samtools
fastqc (optional)
weblogo (optional)
The above software can also be installed using conda, as below:
#Create pipesmrnaseq environment
conda create --name pipesmrnaseq
conda install -n pipesmrnaseq -c bioconda bowtie=1.2.3 #This will automatically install python3.7
conda install -n pipesmrnaseq -c bioconda star
conda install -n pipesmrnaseq -c bioconda bedtools
conda install -n pipesmrnaseq -c bioconda samtools
conda install -n pipesmrnaseq -c bioconda fastqc
conda install -n pipesmrnaseq -c bioconda weblogo
conda install -n pipesmrnaseq -c bioconda git
conda install -n pipesmrnaseq -c conda-forge r-base
conda install -n pipesmrnaseq -c bioconda mirdeep2
#If using mirdeep2 to run miRNA prediction (not included in the pipeline), create an env:
conda create --name mirdeep2
conda install -n mirdeep2 -c bioconda mirdeep2=2.0.1.2
#Create another env for multiqc if you haven't done this before, due to the conflict with pipesmrnaseq:
conda create --name multiqc_env
conda install -n multiqc_env -c bioconda multiqc
#Activate conda env:
conda activate pipesmrnaseq
PipeSmRNAseq.sh is the pipeline script. Other dependencies are in the ./bin folder.
One UCSC tools (from http://hgdownload.soe.ucsc.edu/admin/exe/) is used: faSize. If the binary file the ./bin folder is not working (Execute ./bin/faSize but got errors), please re-download it by choosing the correct version (e.g. linux.x86_64). The faSize for Mac version can be downloaded here: http://hgdownload.soe.ucsc.edu/admin/exe/macOSX.x86_64, and need to be saved in ./bin folder.
Also, for Mac OS, set the pipeline home directory at PipeSmRNAseq.sh manually:
HomeDir="/Users/yusun/Downloads/PipelineHomeDir"
After running this pipeline, we suggest running proTRAC (https://www.smallrnagroup.uni-mainz.de/software.html) to predict piRNA clusters, using mapped small RNAs over 23nt stored in the *final.Over23.${genome}.bed6.sorted file
Here is an example of dm6 genome setup.
1, Download scripts from github to Linux server:
git clone https://github.com/sunyumail93/PipeSmRNAseq.git
mv PipeSmRNAseq PipelineHomeDir
2, Set up index files for genome mapping Use Drosophila melanogaster (dm6) genoem as an example
cd PipelineHomeDir/Ppup/Sequence
3, Set up index files:
mkdir PipelineHomeDir/dm6/Index
cd PipelineHomeDir/dm6/Index
#STAR index, set "--genomeSAindexNbases 12" due to small genome size of dm6:
mkdir STARIndex
STAR --runMode genomeGenerate --genomeDir STARIndex --genomeFastaFiles ../Sequence/dm6.fa --sjdbGTFfile ../Annotation/dm6.RefSeq.reduced.bed12.geneid.gtf --genomeSAindexNbases 12 --sjdbOverhang 100
#bowtie index:
mkdir bowtieIndex
bowtie-build ../Sequence/dm6.fa bowtieIndex/bowtieIndex
#Other index files can be automatically generated when running the pipline.
#Make sure the sequence files for rRNA, tRNA, snRNA, snoRNA, TE are existed in the ./dm6/Sequence folder
4, Compile the C++ program if necessary
The FastqAdapterTimmer binary file was compiled on Linux system. It may need to be compiled again:
g++ FastqAdapterTimmer.cpp -o FastqAdapterTimmer
5, Add executable permissions
chmod +x PipeSmRNAseq.sh
chmod +x ./bin/faSize
PipelineHomeDir/
├── PipeSmRNAseq.sh
├── bin/
└── dm6/
└── Annotation/
├── dm6.RefSeq.reduced.bed12
└── dm6.RefSeq.reduced.bed12.geneid.gtf
└── Index/
├── bowtieIndex
├── Sm_miRNAAllSpeciesIndex/
├── Sm_miRNAIndex/
├── Sm_miRNAMatureIndex/
├── Sm_rRNAIndex/
├── Sm_snoRNAIndex/
├── Sm_snRNAIndex/
├── Sm_TE/
├── Sm_tRNAIndex/
└── STARIndex/
└── Sequence/
├── dm6.ChromInfo.txt
├── dm6.fa
├── dm6.fai
├── dm6.hairpin.fa
├── dm6.mature.fa
├── dm6.rRNA.fa
├── dm6.snoRNA.fa
├── dm6.snRNA.fa
├── dm6.TE.fa
└── dm6.tRNA.fa
└── Ppup/
...
Notes:
1, For the Annotation folder, the GTF file is required for STAR index generation
2, For the Index folder, STAR and bowtie whole genome indexes are big, and it would be better to generate them separately. Other directly bowtie mapping indexes are small and can be generated automatically when running the pipeline.
3, Whole genome sequence and other direct mapping sequences (rRNA, tRNA, snRNA, snoRNA, TE) are in the Sequence folder.
Type the pipeline name, then you will see the manual page:
PipeSmRNAseq.sh
Manual page:
The trimming script also has manual page:
FastqAdapterTimmer
Sometimes we need more trimming at the 5'-end, for example, trim the first 7nt from the left:
FastqSeqClipLeft.py
A regular run using mostly default parameters:
PipeSmRNAseq.sh -i Data.fastq.gz -g dm6
More parameters used:
PipeSmRNAseq.sh -i Data.fastq.gz -g dm6 -p 8
1, Download data
Use a public dataset: GEO SRA: SRR1568355
fastq-dump is part of NCBI SRA Toolkit:
fastq-dump --split-3 SRR1568355
2, Trim adaptor using the trimming script from /bin repository: FastqAdapterTimmer
FastqAdapterTimmer -i SRR1568355.fastq -a AGATCG -o Data.trimmed.fastq
gzip Data.trimmed.fastq
3, Run the small RNAseq pipeline:
PipeSmRNAseq.sh -i Data.trimmed.fastq.gz -g dm6 -p 8
1, Length distribution of small RNA all reads, separated by categories:
2, Nucleotide composition analysis