Small RNA Sequencing · miRNA & sRNA Profiling

Profile regulatory RNAs. Discover miRNA networks.

End-to-end Small RNA Sequencing and analysis for plants, animals and microbes — comprehensive profiling of miRNAs, siRNAs and other short regulatory RNAs for post-transcriptional gene regulation and stress-response studies.

21–24 nt
Key miRNA / siRNA size classes
miRBase
Known miRNA annotation
Novel
miRNA prediction supported
DESeq2
+ edgeR for DE analysis

Small RNA profiling for regulatory biology

Nucleome Informatics offers end-to-end Small RNA Sequencing (sRNA-Seq) and data analysis services for a wide range of organisms including plants, animals and microbes.

Small RNA sequencing enables comprehensive profiling of short regulatory RNA species including microRNAs (miRNAs), small interfering RNAs (siRNAs) and other small non-coding RNAs. This makes sRNA-Seq a primary approach for investigating post-transcriptional gene regulation, stress-response pathways and development-associated regulatory networks.

Our analysis combines rigorous sequencing quality control, adapter trimming, reference-guided or de novo small RNA identification, miRNA annotation against curated databases and differential expression analysis between experimental conditions.

  • Sequencing data quality assessment and preprocessing
  • Small RNA size-class characterisation and abundance profiling
  • Reference genome alignment and small RNA locus identification
  • miRNA annotation against miRBase and organism-specific databases
  • Novel miRNA prediction using established tools
  • Differential expression analysis across experimental conditions
  • Target gene prediction for DE miRNAs
  • Sample-wise and project-level QC and small RNA summaries
  • Secure delivery of raw, intermediate and final analysis files

Multi-Class Profiling

Characterise miRNA, siRNA and other small RNA classes from the same dataset — including 21–24 nt windows typical of plant and animal regulatory RNAs.

miRNA siRNA phasiRNA Other sRNA

Known + Novel miRNAs

Known miRNA identification via miRBase combined with de novo novel miRNA prediction (miRDeep2 / ShortStack) for discovery-focused studies.

Small RNAs, from first principles to the fine print

Pick your level. Each tab builds on the one before, so you can read straight through or jump to the part you need.

Start here if sRNA-Seq is new to you. No background in RNA biology is assumed.

BASIC

What is a small RNA?

A short RNA, roughly 18–30 nucleotides long, that does not code for protein. It is loaded into an Argonaute (AGO) protein, which uses the small RNA as a guide to find matching RNA or DNA and turn that gene down.

BASIC

Why sequence them?

miRNAs fine-tune gene networks in development, stress responses and disease. In plants, siRNAs also guard the genome and shape crop traits. Sequencing counts each molecule one by one and can find small RNAs nobody has described yet.

BASIC

What does the data look like?

Reads are longer than the RNA itself, so the 3′ adapter is read into and trimmed off. What remains is the insert. Its length is a first clue to the class: 21 and 22 nt often mean miRNA or siRNA, 24 nt in plants usually means heterochromatic siRNA.

For readers who want to understand what the tables and plots in a report mean.

INTERMEDIATE

Where they come from

miRNAs are cut from hairpin precursors, by Drosha and Dicer in animals and by DCL1 in plants. siRNAs are cut from double-stranded RNA. Plants have four Dicer-like enzymes with typical products: DCL4 21 nt, DCL2 22 nt, DCL3 24 nt, DCL1 21–22 nt.

INTERMEDIATE

Sorting into Argonautes

The first (5′) nucleotide helps decide which AGO takes the small RNA. In Arabidopsis, AGO1 favours 5′ U, AGO2 and AGO4 favour 5′ A, and AGO5 favours 5′ C (Mi et al., 2008). That is why 5′ base composition by length (Fig. 5) is a useful check.

INTERMEDIATE

Library prep shapes the data

Adapters are ligated to both ends, and the ligases prefer some sequences and structures. Read counts therefore do not equal molecule counts. Comparing the same miRNA across samples made with the same kit is much safer than comparing different miRNAs to each other (Hafner et al., 2011; Giraldez et al., 2018).

INTERMEDIATE

What makes a real miRNA

Matching a database entry is easy. Calling a new miRNA is harder: the reads must come from a hairpin, with a precise 5′ end and a partner (miRNA*) strand offset by two nucleotides. Plant criteria were tightened in 2018 (Axtell and Meyers) because many earlier annotations were siRNAs.

INTERMEDIATE

Counting and comparing

Counts depend on sequencing depth, so they are scaled before comparing samples. Simple scaling by total reads can mislead when a few miRNAs dominate the library. Replicates are what allow statistics at all: with none, a fold-change is only a number.

INTERMEDIATE

From miRNA to target

Target prediction lists genes a miRNA could regulate based on sequence pairing. It is a hypothesis list. In plants, degradome sequencing can confirm cleavage; in animals, extra evidence is usually needed.

For specialists who want to know where the method is strong, and where it needs care.

ADVANCED

IsomiRs

One locus produces a family of variants that differ at the 5′ or 3′ end. A shift at the 5′ end moves the seed and changes the target set. Many 3′ differences are templated trimming or tailing (for example added U or A), but some can also arise during library preparation or sequencing, so they need replicate support.

ADVANCED

Plant versus animal targeting

Animal miRNAs need only about 6–8 nt of seed pairing, so one miRNA can have hundreds of targets. Plant miRNAs usually pair along most of their length and direct cleavage (Rhoades et al., 2002; Llave et al., 2002). The same prediction logic therefore does not transfer between kingdoms.

ADVANCED

Phased siRNAs

A 22-nt miRNA (or a two-hit trigger, as with miR390 and TAS3) can cut a transcript and recruit RDR6. DCL4 then slices the double-stranded product in 21-nt steps from the cut site, giving a recognisable register (Allen et al., 2005; Axtell et al., 2006). Reproductive phasiRNAs are sometimes called the piRNAs of plants (Liu et al., 2020).

ADVANCED

24-nt siRNAs and RdDM

Most 24-nt reads in plants come from repeats and transposons. Pol IV, RDR2 and DCL3 make them, AGO4 carries them, and Pol V and DRM2 methylate the DNA (Matzke and Mosher, 2014). Because the sources are repetitive, many reads map to several places.

ADVANCED

Composition effects

A sequencing library is a fraction of 100%. If one abundant miRNA rises, all others appear to fall. Median-based normalisation (DESeq2, TMM) fixes this when most miRNAs are unchanged, but not when a large share change in the same direction (see Fig. 12).

ADVANCED

Multi-mapping and families

Reads from miRNA family members, rRNA-derived fragments and repeats often match several loci. Strategies include unique-only counting, fractional counting, and weighting by nearby unique reads, which is the approach ShortStack takes. Family-level counts are more robust than locus-level counts when members are near-identical.

miRNA
~21–22 nt RNA cut from a hairpin precursor; guides repression of target RNAs.
siRNA
Small RNA cut from long double-stranded RNA; 21–24 nt in plants.
phasiRNA
siRNA made in a regular 21- or 24-nt register after a miRNA cut.
Seed
Nucleotides 2–8 of a miRNA, the key pairing region in animals.
AGO
Argonaute protein that carries the small RNA and does the silencing.
miRNA*
The partner strand of the duplex, usually far less abundant than the miRNA.
isomiR
A sequence variant of a miRNA, differing at its ends.
Degradome
Sequencing of cleaved RNA ends; shows where a miRNA has cut.

How small RNAs are made, and how they show up in your data

Schematics explain the biology; the plots below use simulated data to show what the sequencing signal looks like and what to check. None of the data here comes from a client project.

Figure 1 Biogenesis of miRNAs in animals and plants schematic
AnimalsNUCLEUSpri-miRNA (RNA Pol II)Drosha + DGCR8 cut the hairpinpre-miRNA, about 60–70 ntCYTOPLASMExportin-5Dicer + TRBP remove the loop~22-nt miRNA duplexAGO loading; passenger strand discardedRISC: seed pairing (nt 2–8) to the 3′ UTRtranslational repression and mRNA decayPlantsNUCLEUS (all processing happens here)pri-miRNA (RNA Pol II)DCL1 + HYL1 + SE: two cuts releasea miRNA/miRNA* duplex (21–22 nt)HEN1 adds 2′-O-methyl to both 3′ endsexport to cytoplasm (HASTY and others)CYTOPLASMAGO1 loadingRISC: near-complete pairing, AGO1 cleavage(translational repression also occurs)
Figure 1 | Same idea, different machinery. Animals cut in two places: Drosha in the nucleus, Dicer in the cytoplasm. Plants do both cuts with DCL1 inside the nucleus, then methylate the duplex with HEN1 (Bartel, 2018; Rogers and Chen, 2013).
Figure 2 A real miRNA locus has a precise read stack simulated
Figure 2 | Precision is the signature. Left: reads from a miRNA hairpin start at the same position (the miRNA) with a small partner peak from the other arm (miRNA*). Right: an siRNA-type locus produces ragged, widely spread 5′ ends. This is the logic behind hairpin-and-read-stack tools such as miRDeep2 and ShortStack, and behind the plant criteria of Axtell and Meyers (2018).
Figure 3 How miRNAs find their targets schematic
Animal: seed pairingmiRNA 5′12810112122target mRNA 3′ UTRseed (nt 2–8)Only about 6–8 nt need to pair, so one miRNA canregulate hundreds of transcriptsTypical outcome: repression, deadenylation and decayPlant: extensive complementaritymiRNA 5′128101121target transcriptAGO cleaves between nt 10 and 11Most positions pair, with only a few mismatches allowed.Fewer targets per miRNA, often transcription factorsTypical outcome: cleavage, detectable by degradome-seq
Figure 3 | Seed pairing versus cleavage. Orange circles mark the seed on the left and the cleavage position on the right. Because the rules differ, target tools are built separately for animals and plants (Lewis et al., 2005; Dai et al., 2018).
Figure 4 Read-length profile and the size filter interactive · simulated
Figure 4 | The length profile is a fingerprint. A plant sample (left) typically shows peaks at 21 and 24 nt; an animal tissue sample (right) peaks at 22 nt, with a shoulder of longer tRNA- and rRNA-derived fragments. Green bars are inside the length window you choose; grey bars are discarded. Window choice is part of the study design.
Figure 5 5′ nucleotide by read length illustrative
Figure 5 | AGO sorting leaves a mark. 21-nt reads lean towards 5′ U (AGO1-type miRNAs); 24-nt reads lean strongly towards 5′ A (AGO4-type). A 24-nt peak with a different 5′ profile would be a flag for contamination or a library problem.
Figure 6 Phasing: reads in a 21-nt register simulated
Figure 6 | A regular ladder of reads. After a miRNA cut at position 20, 21-nt products appear every 21 nt (left, green). Right: share of reads falling into each of the 21 possible registers. A phased locus concentrates in one register; an unphased locus spreads evenly.
Figure 7 RNA-directed DNA methylation (RdDM) schematic
Pol IVmakes a shorttranscriptRDR2copies it intodouble-stranded RNADCL3cuts dsRNA into24-nt duplexesHEN12′-O-methylatesthe 3′ endsAGO4 / AGO6load the 24-nt siRNA(5′ A preferred)Pol Vmakes a scaffold RNA thatthe siRNA base-pairs withDRM2is recruited and addscytosine methylationChromatinmethylated DNA, histone markssilenced repeats and TEs
Figure 7 | Where 24-nt reads come from. The canonical RdDM pathway makes 24-nt siRNAs that guide DNA methylation to repeats and transposons (Matzke and Mosher, 2014). The pathway runs top row left to right, then bottom row right to left.

Built for small RNA biology

From size-class profiling through target prediction — a complete sRNA analysis toolkit.

Size-Class Profiling

Read length distribution analysis to characterise miRNA, siRNA and other small RNA species.

miRBase Annotation

Known miRNA identification and quantification against the current miRBase release for the relevant organism.

Novel miRNA Prediction

De novo prediction of novel miRNA candidates using tools such as miRDeep2 or ShortStack.

Differential Expression

Robust statistical DE analysis using DESeq2 and/or edgeR for miRNA and small RNA comparisons.

Target Gene Prediction

Identification of putative target genes for DE miRNAs to support downstream functional interpretation.

Broad Organism Support

Analysis supported for human, animal and plant sRNA datasets, including non-model organisms.

From raw reads to annotated sRNA profiles

Each sample passes through a quality-controlled analysis workflow designed to move from raw sequencing data to annotated, differentially expressed small RNA profiles.

Workflow A QC-gated view of the workflow schematic
STEP 01

QC and preprocessing

Check base quality, locate the 3′ adapter, trim it and remove low-quality or too-short inserts.

insert-length peak present
FASTQQC report
STEP 02

Size-class characterisation

Plot the length distribution and 5′ nucleotide by length to see which classes are in the sample.

21 / 22 / 24 nt profile as expected
length plot
STEP 03

Reference alignment

Map trimmed reads to the genome with an aligner suited to short reads and a stated multi-mapper policy.

mapping rate consistent across samples
BAM
STEP 04

miRNA annotation

Match reads to miRBase and organism-specific annotation, then predict novel miRNAs from hairpin and read-stack patterns.

precision criteria met
knownnovel
STEP 05

Quantification

Count reads per miRNA and per small RNA class, producing matrices for every sample.

replicates correlate
miRNA matrixclass matrix
STEP 06

Differential expression

Test for differences between groups with DESeq2 or edgeR, controlling the false discovery rate.

PCA groups by condition
DESeq2edgeR
STEP 07

Target prediction

Predict candidate target genes for the differentially expressed miRNAs.

scoring cut-off documented
target table
STEP 08

Reporting and delivery

Summary report with QC and result figures, with files shared through secure delivery.

checks signed off
reportsecure link
Workflow | One page for the whole journey. Each gate is a question we ask of the data before moving on. Tools and thresholds are examples of their kind and are finalised per project.
01

Data QC & Pre-processing

Quality assessment, adapter trimming and low-quality read removal.

02

Size-Class Characterisation

Read length distributions analysed for miRNA and siRNA size windows.

03

Reference Alignment

Filtered reads aligned to the reference genome to map small RNA loci.

04

miRNA Annotation

Known miRNAs via miRBase; novel candidates predicted from structure and abundance.

05

Quantification — miRNA-level and small RNA class-level expression matrices for DE analysis.

06

Differential Expression — DESeq2 and/or edgeR to identify significant miRNA / sRNA changes.

07

Target Prediction & Reporting — Putative targets for DE miRNAs; QC, profiles and DE results delivered.

Trust the library before you trust the biology

Small RNA libraries fail in characteristic ways: degraded RNA, adapter dimers, dominant abundant species. These plots use simulated data to show what to look for.

Figure 8 What the reads are made of simulated
Figure 8 | Read-class composition per sample. Eight simulated libraries. Sample T3 has a high rRNA share and a low miRNA share, the typical pattern of degraded input RNA. Such samples are flagged before any comparison is made, and the decision to keep or exclude them is agreed with the client.
Figure 9 Is sequencing deep enough? simulated
Figure 9 | Saturation of detected miRNAs. Expected number of miRNAs with at least 5 reads as miRNA-assigned reads increase, for three levels of complexity. Curves flatten once most expressed miRNAs are seen; low-abundance miRNAs keep emerging much longer.
Figure 10 A few families dominate simulated
Figure 10 | Rank-abundance of miRNA families. Bars show each family's share of miRNA reads; the line is the cumulative share. In real tissues a handful of families often carry most reads, which is why normalisation must cope with composition (Fig. 12).
Figure 11 IsomiRs of one miRNA illustrative
Figure 11 | Variants around a canonical sequence. Reads for one miRNA-like example. Orange letters are added bases, and shortened ends are shown by missing letters. Counting at miRNA level groups these variants; a shift at the 5′ end changes the seed and is worth examining separately (Neilsen et al., 2012).

Finding the miRNAs that really change

Statistics ask which miRNAs differ between groups once biological variation is taken into account. Effect size and confidence should be read together, with false discovery rate control across everything tested.

Figure 12 Normalisation and the composition problem interactive · simulated
Figure 12 | Why total-count scaling can mislead. 100 simulated miRNAs, of which the chosen share is raised by the chosen fold-change. Dots are the log2 fold-change of miRNAs that truly did not change (grey) and ones that did (green). With scaling by total reads, unchanged miRNAs appear to fall. Median-based scaling (the idea behind DESeq2 and TMM) corrects this while fewer than half of the miRNAs change, but fails when more than half do (Love et al., 2014; Robinson and Oshlack, 2010).
Figure 13 Sample relationships (PCA) simulated
Figure 13 | Do replicates group together? Principal components of log-scaled miRNA counts for five control and five treated samples. Tight groups that separate on PC1 support the design; an outlier or a split inside a group points to batch or quality issues to resolve first.
Figure 14 Volcano plot of differential miRNAs interactive · simulated
Figure 14 | Effect size versus confidence. Each point is a simulated miRNA. Green points are up in treated, orange are down, both passing the two thresholds you set. Adjusted p-values use the Benjamini–Hochberg procedure (1995).
Figure 15 Heatmap of the top differential miRNAs simulated
Figure 15 | Row-scaled expression. Each row is a miRNA family, each column a sample; colour shows how far above (orange) or below (teal) that miRNA's own average a sample lies. Families are named generically; real reports use the exact miRBase or novel IDs.
Table 1 DESeq2 or edgeR?
MethodHow it worksNotes for sRNA-Seq
DESeq2Negative-binomial model, median-of-ratios size factors, shrunken fold-changesShrinkage stabilises fold-changes of noisy low-count miRNAs
edgeRNegative-binomial model, TMM scaling, empirical-Bayes dispersionVery similar in practice; flexible designs and quasi-likelihood tests
BothCount-based, model replicatesAgreement between the two is a useful sanity check; both assume most miRNAs do not change
Replicates matter more than tools. With one sample per group, no method can separate treatment from sample-to-sample noise. The service uses both DESeq2 and edgeR as described above; results that hold in both are the most dependable.
Also check the filter. Very low-count miRNAs are removed before testing, which improves power; the filter is stated in the report.

From differential miRNAs to candidate targets

A prediction is a ranked list of hypotheses. Thresholds trade missed targets against false ones.

Figure 16 Choosing a score cut-off interactive · illustrative
Figure 16 | A cut-off is a trade-off. Simulated score distributions for genuine and background sites, where lower scores mean better pairing. Tightening the cut-off removes background but also loses real targets. The scoring scheme differs between tools, so the cut-off used is recorded with the results.
Figure 17 What the targets have in common simulated
Figure 17 | Functional enrichment of predicted targets. Dot position shows the share of targets in a category, dot size the gene count, and colour the adjusted p-value. Enrichment of predicted targets supports interpretation but does not on its own prove regulation (Yu et al., 2012).
From prediction to proof. In plants, degradome (PARE) sequencing finds cleavage products and can confirm a predicted cut site (German et al., 2008; Addo-Quaye et al., 2009). In animals, reporter assays, AGO-binding data or paired miRNA and mRNA profiles are the usual next steps.

Structured outputs for sRNA research

At project completion, clients receive a structured set of sequencing, analysis and reporting outputs via secure shared storage.

Raw & Processed Data

FASTQ files — raw and appropriately trimmed sequence data.

QC Reports

Sample-level QC summaries including read length distributions.

Alignment Files

Small RNA read alignments to the reference genome.

Expression Matrices

miRNA and small RNA class-level count and expression matrices.

DE Results

Per-comparison tables of DE miRNAs with statistics and fold changes.

Annotated Outputs

Known and novel miRNA annotations with target predictions for DE candidates.

Summary Report

Sample-wise QC, alignment, quantification and DE analysis statistics.

Secure Delivery

Project files via secure shared storage or cloud-based access.

What we accept

Flexible input options for sequencing-plus-analysis or analysis-only projects.

Small RNA-Enriched Total RNA

We accept total RNA enriched for small RNA species from plant, animal, cell or other biological material, as well as pre-generated sequencing data where analysis-only services are required.

Project-specific requirements for RNA quantity, concentration, integrity, sequencing depth and read length are finalised according to the study design and organism. Small RNA library preparation typically requires input total RNA of sufficient quality and quantity to ensure small RNA representation.

Contact us for a project-specific sample submission checklist and recommended sequencing depth.

~21–24 nt

Plant · Animal · Cells
or pre-generated FASTQ

Enriched, not just extractedSmall RNA libraries work best from RNA that retains the 18–30 nt fraction. Methods that discard short RNAs during clean-up reduce what can be seen.
Replicates firstStatistics need biological replicates. Typical designs use at least three per group; we agree the number with you for the question being asked.
Tell us the kitFor existing FASTQ files, the adapter sequences, any random bases in the adapters and the read length decide how reads are trimmed. Please share the library kit name if you know it.
Name the referenceGive us the organism and genome version. Annotation quality differs a lot between species, and plants with polyploid or repeat-rich genomes need extra care.

Proven across organisms & sRNA classes

Production workflows applied to miRNA profiling and plant siRNA characterisation.

miRNA Profiling in Oryza sativa

Rice · Multi-condition · miRBase + DESeq2

Mean raw-data Q30>90%
Reads mapping to known miRNAsSample-dependent
Known miRNAs quantifiedOrganism-dependent
Novel miRNA candidatesReported with confidence
DE analysisDESeq2

Multi-condition sRNA-Seq study in rice processed end-to-end: QC, reference alignment, miRBase-based annotation, novel miRNA prediction and DESeq2-based differential expression analysis.

Plant siRNA Profiling

24 nt siRNA · phasiRNA · Heterochromatic siRNA

Size class focus24 nt siRNA
phasiRNA supportYes
Heterochromatic siRNAYes
Epigenetic relevancePlant systems
WorkflowSame validated pipeline

For studies requiring characterisation beyond miRNAs, our workflow supports profiling of 24 nt siRNA populations including phased siRNAs (phasiRNAs) and heterochromatic siRNAs relevant to epigenetic regulation in plant systems.

What small RNA-Seq can tell you, and where the field is heading

Annotation standards, library chemistry and the biology of small RNA classes are all moving. Here is the landscape, so you can place your own project in it.

Table 2 Annotation resources compared published data
ResourceWhat it holdsTake-away
miRBase (v22)More than 38,000 hairpin precursors from 271 organisms, including plantsBroadest coverage; some entries are doubtful and have been questioned (Kozomara et al., 2019; Axtell and Meyers, 2018)
MirGeneDB 3.0More than 20,000 curated miRNA genes in 1,743 families from 114 metazoan speciesStrict curation and uniform nomenclature; animals only (preprint, 2024)
Organism-specific setsCurated lists for individual plant or animal speciesOften the best fit for a given genome; quality varies by species
Why this matters. Reporting which database and release was used is part of reproducibility. A miRNA that is "known" in one set can be absent or reclassified in another.
Table 3 Who needs which analysis?
QuestionApproachStatus here
Which miRNAs differ?Annotation, counting, DESeq2/edgeRIncluded
Are there new miRNAs?Hairpin and read-stack predictionIncluded
What do they regulate?Target prediction, optional degradomePrediction included
Which classes are present?Size-class and 5′ nucleotide profilingIncluded
Phased loci, isomiRs, tRFs, piRNAsSpecialised, project-specific analysesDiscuss with us
ISOMIRS

Variants carry information

Added U or A at the 3′ end, or 5′ shifts, can reflect real biology. Replicate support and a chemistry-aware view of library bias are needed before interpreting single variants (Neilsen et al., 2012; Giraldez et al., 2018).

TRFS · PIRNAS

Beyond miRNA

Fragments of tRNAs and, in animal germlines, PIWI-bound piRNAs make up a large share of some libraries. Several tRNA fragments associate with Argonaute proteins (Kumar et al., 2014), but their annotation is still being standardised.

LIBRARY CHEMISTRY

Randomised adapter ends

Adapters with random bases at the ligation end reduce sequence-specific ligation bias in comparative studies (Fuchs et al., 2015; Giraldez et al., 2018). Tell us if your libraries use them so reads can be handled correctly.

ABSOLUTE LEVELS

Spike-ins and global shifts

When many miRNAs shift together, relative normalisation can hide the change. Synthetic spike-in oligos added at extraction anchor the measurement to a fixed amount (Lovén et al., 2012).

VALIDATION

Degradome sequencing

Sequencing the 5′ ends of cleaved RNA pinpoints where a miRNA has cut, giving direct evidence for plant targets (German et al., 2008).

SPECIES ORIGIN

Contamination checks

Tools that infer the organism of origin from miRNA content help catch mislabelled or cross-contaminated samples (Kang et al., 2018).

Scope note. Nucleome's small RNA-Seq service covers the workflow and deliverables described on this page, from QC and preprocessing through size-class characterisation, alignment, miRBase and novel miRNA annotation, differential expression with DESeq2 and edgeR, and target prediction. Phased loci, isomiR-level, tRF and piRNA analyses, spike-in designs and degradome work shown here place the service in its scientific context; contact our team to discuss project-specific extensions.

Service highlights

Support for research, agricultural, biomedical and regulatory biology sRNA-Seq projects.

End-to-End sRNA-Seq

From raw sequencing data through QC, alignment, annotation, DE analysis and target prediction.

Comprehensive Annotation

Known miRNA identification via miRBase combined with de novo novel miRNA prediction.

Multi-Class Profiling

Support for miRNA, siRNA and other small RNA class characterisation from the same dataset.

Established Statistics

DE analysis using DESeq2 and edgeR for robust, reproducible results.

Target Prediction

Putative target gene identification for DE miRNAs to support functional interpretation.

Broad Organism Support

Reference-based workflows for human, animal and plant sRNA datasets.

Frequently asked, in plain language

From the basics to the finer points of study design and statistics.

basicWhat is small RNA sequencing?
It reads the short RNAs, roughly 18–30 nucleotides long, that guide gene silencing. It tells you which small RNAs are present and how much of each, including miRNAs, siRNAs and related classes.
basicWhat samples can I send?
Total RNA enriched for small RNAs from plant, animal, cell or other material. You can also send existing FASTQ files for analysis only.
basicWhat will I receive?
FASTQ files, QC reports with read-length distributions, alignment files, expression matrices, differential expression results, annotated outputs with target predictions, a summary report, and secure delivery.
intermediateWhy do the length peaks matter?
The insert length points to the class. In plants, peaks at 21 and 22 nt suggest miRNAs and some siRNAs, and a large 24-nt peak suggests heterochromatic siRNAs. In animals, a sharp peak near 22 nt suggests miRNAs. A missing or shifted peak often signals a library or RNA problem.
intermediateHow many replicates and how much depth do I need?
It depends on the organism, how many miRNAs you hope to detect and how large the effect is. Replicates matter more than depth: we typically recommend at least three per group and agree depth and design with you per study.
intermediateIs miRBase enough?
It is the standard starting point, and the service uses it together with organism-specific annotation. Some entries are doubtful, so novel and known calls are reported separately, with the criteria stated.
intermediateHow are novel miRNAs predicted?
From the genome: reads must come from a hairpin-forming locus with a precise 5′ end and a partner strand, as in tools such as miRDeep2 and ShortStack. Plant calls follow stricter published criteria (Axtell and Meyers, 2018).
advancedWhy do my miRNA counts not match qPCR?
Adapter ligation favours some sequences, so read counts are not proportional to molecules. Relative changes of the same miRNA between samples are far more reliable than comparisons between different miRNAs, and a qPCR check of key hits is good practice.
advancedWhat if most miRNAs change in my experiment?
Median-based normalisation assumes most miRNAs are unchanged. If a large share move together, relative scaling can hide or exaggerate changes. Tell us at the design stage; spike-in standards are one way to anchor the measurement.
advancedCan predicted targets be trusted?
They are hypotheses. Animal predictions are many and often weakly supported; plant predictions are fewer and can be confirmed with degradome data. We document the tool and cut-off so the list can be reproduced.

Key literature & resources

Selected sources behind the methods and reference values on this page. Figures labelled "simulated", "schematic" or "illustrative" are generated for this page, not client data.

  1. Bartel DP. Metazoan MicroRNAs. Cell 2018;173:20–51.
  2. Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell 2009;136:215–233.
  3. Kim VN, Han J, Siomi MC. Biogenesis of small RNAs in animals. Nature Reviews Molecular Cell Biology 2009;10:126–139.
  4. Rogers K, Chen X. Biogenesis, turnover, and mode of action of plant microRNAs. Plant Cell 2013;25:2383–2399.
  5. Bologna NG, Voinnet O. The diversity, biogenesis, and activities of endogenous silencing small RNAs in Arabidopsis. Annual Review of Plant Biology 2014;65:473–503.
  6. Axtell MJ. Classification and comparison of small RNAs from plants. Annual Review of Plant Biology 2013;64:137–159.
  7. Axtell MJ, Meyers BC. Revisiting criteria for plant microRNA annotation in the era of big data. Plant Cell 2018;30:272–284. Record
  8. Yu B, et al. Methylation as a crucial step in plant microRNA biogenesis. Science 2005;307:932–935.
  9. Mi S, et al. Sorting of small RNAs into Arabidopsis Argonaute complexes is directed by the 5′ terminal nucleotide. Cell 2008;133:116–127.
  10. Kozomara A, Birgaoanu M, Griffiths-Jones S. miRBase: from microRNA sequences to function. Nucleic Acids Research 2019;47:D155–D162.
  11. Fromm B, et al. MirGeneDB 2.1: toward a complete sampling of all major animal phyla. Nucleic Acids Research 2022;50:D204–D210.
  12. MirGeneDB 3.0: Improved taxonomic sampling, uniform nomenclature of novel conserved microRNA families, and updated covariance models. bioRxiv 2024 (preprint). Record
  13. Fromm B, et al. A uniform system for the annotation of vertebrate microRNA genes and the evolution of the human microRNAome. Annual Review of Genetics 2015;49:213–242.
  14. Friedländer MR, et al. Discovering microRNAs from deep sequencing data using miRDeep2. Nucleic Acids Research 2012;40:37–52.
  15. Axtell MJ. ShortStack: comprehensive annotation and quantification of small RNA genes. RNA 2013;19:740–751. Record
  16. Johnson NR, Yeoh JM, Coruh C, Axtell MJ. Improved placement of multi-mapping small RNAs. G3 2016;6:2103–2111.
  17. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.journal 2011;17:10–12.
  18. Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biology 2009;10:R25.
  19. Hafner M, et al. RNA-ligase-dependent biases in miRNA representation in deep-sequenced small RNA cDNA libraries. RNA 2011;17:1697–1712.
  20. Fuchs RT, et al. Bias in ligation-based small RNA sequencing library construction is determined by adaptor and RNA structure. PLoS ONE 2015;10:e0126049.
  21. Giraldez MD, et al. Comprehensive multi-center assessment of small RNA-seq methods for quantitative miRNA profiling. Nature Biotechnology 2018;36:746–757.
  22. Dard-Dascot C, et al. Systematic comparison of small RNA library preparation protocols for next-generation sequencing. BMC Genomics 2018;19:118. Record
  23. Neilsen CT, Goodall GJ, Bracken CP. IsomiRs – the overlooked repertoire in the dynamic microRNAome. Trends in Genetics 2012;28:544–549.
  24. Allen E, Xie Z, Gustafson AM, Carrington JC. microRNA-directed phasing during trans-acting siRNA biogenesis in plants. Cell 2005;121:207–221.
  25. Axtell MJ, Jan C, Rajagopalan R, Bartel DP. A two-hit trigger for siRNA biogenesis in plants. Cell 2006;127:565–577.
  26. Cuperus JT, et al. Unique functionality of 22-nt miRNAs in triggering RDR6-dependent siRNA biogenesis from target transcripts in Arabidopsis. Nature Structural & Molecular Biology 2010;17:997–1003.
  27. Liu Y, Teng C, Xia R, Meyers BC. PhasiRNAs in plants: their biogenesis, genic sources, and roles in stress responses, development, and reproduction. Plant Cell 2020;32:3059–3080.
  28. Matzke MA, Mosher RA. RNA-directed DNA methylation: an epigenetic pathway of increasing complexity. Nature Reviews Genetics 2014;15:394–408.
  29. Kumar P, et al. Meta-analysis of tRNA derived RNA fragments reveals that they are evolutionarily conserved and associate with AGO proteins to recognize specific RNA targets. BMC Biology 2014;12:78.
  30. Ozata DM, et al. PIWI-interacting RNAs: small RNAs with big functions. Nature Reviews Genetics 2019;20:89–108.
  31. Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 2005;120:15–20.
  32. Agarwal V, et al. Predicting effective microRNA target sites in mammalian mRNAs. eLife 2015;4:e05005.
  33. Rhoades MW, et al. Prediction of plant microRNA targets. Cell 2002;110:513–520.
  34. Llave C, Xie Z, Kasschau KD, Carrington JC. Cleavage of Scarecrow-like mRNA targets directed by a class of Arabidopsis miRNA. Science 2002;297:2053–2056.
  35. Dai X, Zhuang Z, Zhao PX. psRNATarget: a plant small RNA target analysis server (2017 release). Nucleic Acids Research 2018;46:W49–W54.
  36. German MA, et al. Global identification of microRNA–target RNA pairs by parallel analysis of RNA ends. Nature Biotechnology 2008;26:941–946.
  37. Addo-Quaye C, Miller W, Axtell MJ. CleaveLand: a pipeline for using degradome data to find cleaved small RNA targets. Bioinformatics 2009;25:130–131.
  38. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 2014;15:550.
  39. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 2010;26:139–140.
  40. Robinson MD, Oshlack A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biology 2010;11:R25.
  41. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society B 1995;57:289–300.
  42. Lovén J, et al. Revisiting global gene expression analysis. Cell 2012;151:476–482.
  43. Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16:284–287.
  44. Kang W, et al. miRTrace reveals the organismal origins of microRNA sequencing data. Genome Biology 2018;19:213.