Small RNA-seq Pipeline - Usage Guide

Overview

Complete workflow from small RNA-seq FASTQ to differential miRNA expression and target prediction, with the small-RNA-specific decisions surfaced at each step. The pipeline trims kit-specific adapters, quantifies known miRNAs and isomiRs (miRge3) or discovers novel miRNAs (miRDeep2), runs compositionally-aware differential expression (DESeq2 on raw counts), and predicts targets filtered by expression evidence. When abundant non-miRNA classes appear, it branches to tRF/piRNA profiling.

Prerequisites

conda install -c bioconda cutadapt mirdeep2 bowtie viennarna miranda
pip install mirge3 umi_tools
# BiocManager::install(c('DESeq2', 'apeglm'))

Quick Start

• "Analyze my small RNA-seq from FASTQ to differential miRNAs and targets"
• "Quantify known miRNAs with miRge3, then run DESeq2 on the raw counts"
• "Discover novel miRNAs with miRDeep2 and validate the candidates"
• "My library is mostly tRFs, not miRNAs - what should I run?"

Example Prompts

Full Pipeline

"Run the complete small RNA-seq pipeline from trimming to target prediction"

"Find DE miRNAs between my conditions and predict their targets"

Specific Steps

"Just quantify known miRNAs and isomiRs with miRge3"

"Predict targets for my DE miRNAs and intersect with my mRNA-seq"

"Profile the tRFs and piRNAs instead of miRNAs"

Pipeline-level decisions

"Which normalization should I use - my library is dominated by a few miRNAs?"

"My samples are plasma/serum - how do I normalize when there's no stable endogenous miRNA?"

"Should I quantify known miRNAs (miRge3) or discover novel ones (miRDeep2)?"

"My size factors look skewed and lots of miRNAs are 'down' - is that real?"

What the Agent Will Do

1. Trim the kit adapter (discarding untrimmed reads) and handle any 4N spacer or UMI
2. Quantify known miRNAs/isomiRs with miRge3, or discover novel miRNAs with miRDeep2 when needed
3. Check the RNA-class composition and branch to tRF/piRNA profiling if non-miRNA classes dominate
4. Run DESeq2 on raw counts, inspecting size factors for compositional distortion
5. Predict targets and filter them by anti-correlated mRNA DE and validated databases

Tips

• The 3' adapter is on every real read, so --discard-untrimmed is correct; strip NEXTflex 4N after trimming, extract QIAseq UMIs before/after alignment
• The length histogram is QC: a 21-23 nt peak is a good miRNA library; a 26-32 nt peak is piRNA; a 30+ nt smear is degradation or contamination
• Ligation bias makes absolute cross-miRNA abundance untrustworthy; compare the same miRNA across samples, never across kits
• Feed RAW counts to DESeq2 (RPM is display only) and inspect size factors - a few dominant miRNAs distort normalization
• A miRDeep2 score is a hypothesis; pick a cutoff from survey.pl signal-to-noise and validate novel calls
• A predicted target is a hypothesis; filter by anti-correlated mRNA DE before trusting it
• If non-miRNA classes dominate, this is a tRF/piRNA analysis - switch tools accordingly

Related Skills

• small-rna-seq/smrna-preprocessing - Adapter, UMI, and 4N handling
• small-rna-seq/mirdeep2-analysis - Novel miRNA discovery
• small-rna-seq/mirge3-analysis - Known-miRNA and isomiR quantification
• small-rna-seq/differential-mirna - Compositionally-aware DE
• small-rna-seq/target-prediction - Seed prediction filtered by expression
• small-rna-seq/trf-pirna-profiling - tRF and piRNA profiling
• workflow-management/nf-core-pipelines - Run nf-core/smrnaseq as a curated, reproducible pipeline instead of chaining the steps by hand