Time-Series Gene Set Enrichment Analysis — Dynamic Pathway Enrichment Across Time Points
Time-Series Gene Set Enrichment Analysis · Also known as: longitudinal GSEA, dynamic GSEA, time-course GSEA, TS-GSEA
Time-series gene set enrichment analysis (TS-GSEA) extends the classical GSEA framework to detect biologically coordinated gene sets — pathways, gene ontology terms, or curated signatures — whose collective expression changes meaningfully over time. Rather than comparing two snapshots, it models the full temporal trajectory of gene expression to identify which functional programs are activated, suppressed, or dynamically remodelled during a biological process such as development, treatment response, or disease progression.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use TS-GSEA when you have RNA-seq data from an ordered time-course experiment — developmental stages, treatment time points, circadian cycles, or longitudinal patient sampling — and your question is which biological pathways or programs are coordinately regulated over time, not just at a single contrast. It is especially powerful for detecting gradual, progressive, or oscillatory pathway-level trends that are invisible when collapsing the data into a single fold-change comparison. Do NOT use TS-GSEA when you have only two time points (use standard GSEA instead), when sample sizes per time point are very small (n < 3 per point, which makes temporal modelling unreliable), or when the primary interest is in specific individual genes rather than gene sets. It is also inappropriate for cross-sectional datasets without genuine temporal ordering.
Strengths & limitations
- Captures coordinated gene set dynamics across the full time course rather than at a single comparison, increasing biological sensitivity.
- Reduces the multiple testing burden by operating at the gene set level rather than reporting thousands of individual differentially expressed genes.
- Leverages well-curated databases (MSigDB, KEGG, Reactome) to provide immediately interpretable biological context.
- Compatible with standard GSEA software (GSEA desktop, fgsea R package) when paired with an appropriate temporal ranking step.
- Handles complex temporal patterns — monotonic trends, impulse responses, oscillations — depending on the choice of temporal model.
- Requires at least three ordered time points; studies with only two time points gain no benefit from temporal modelling.
- Temporal model selection (polynomial degree, spline knots, impulse parameters) influences results and requires biological justification.
- Low replication per time point (n < 3) severely limits the power of the temporal regression, making downstream enrichment scores unreliable.
- Results are sensitive to the choice of gene set database and the version used; databases evolve over time, complicating reproducibility.
- Standard GSEA permutation schemes (label permutation) may not correctly preserve temporal autocorrelation structure; specialised permutation strategies are needed.
Frequently asked
Can I use standard GSEA directly on time-series data without a temporal model?
Only if you restrict the comparison to two time points (e.g., baseline vs. final). If you have more than two time points, running GSEA on a single pairwise contrast discards all intermediate information. The correct approach is to first derive a temporal summary statistic per gene from a regression or equivalent model, then use that statistic to rank genes for enrichment scoring.
Which R packages are commonly used for TS-GSEA?
A typical pipeline combines maSigPro or ImpulseDE2 (for temporal differential expression and gene ranking) with fgsea or the GSEA desktop application (for enrichment scoring). The clusterProfiler package provides convenient wrappers for GSEA-style analysis including KEGG and GO databases. For circadian data, JTK_CYCLE or RAIN are used to extract temporal statistics before enrichment testing.
How many time points and replicates do I need?
A minimum of three ordered time points is necessary for temporal modelling to be meaningful. Polynomial regression approaches in maSigPro require at least as many degrees of freedom as the polynomial order plus residual degrees per time point. Practically, three to five time points with two to three biological replicates per point is workable, but statistical power improves markedly with four or more time points and three or more replicates.
What is the difference between TS-GSEA and standard GSEA on a single time-point comparison?
Standard GSEA operates on a ranked gene list derived from one contrast (e.g., treated vs. untreated at 24 h). TS-GSEA derives the ranking from a statistic that summarises the gene's behaviour across the entire time course — such as the overall F-statistic for temporal change or a weighted sum of sequential fold changes. This allows detection of pathway-level dynamics that are not captured by any single time-point comparison alone.
How should I choose the gene set database?
For general pathway analysis, the Hallmark or C2 curated gene sets in MSigDB are widely used and well-validated. For specific biological questions, KEGG and Reactome provide pathway-level resolution with mechanistic annotations. GO Biological Process is appropriate when functional terms rather than curated pathways are of interest. Always record the database version used, as gene set membership changes between releases and affects reproducibility.
Sources
- Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette, M. A., Paulovich, A., Pomeroy, S. L., Golub, T. R., Lander, E. S., & Mesirov, J. P. (2005). Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences, 102(43), 15545–15550. DOI: 10.1073/pnas.0506580102 ↗
- Nueda, M. J., Tarazona, S., & Conesa, A. (2014). Next maSigPro: updating maSigPro bioconductor package for RNA-seq time series. Bioinformatics, 30(18), 2598–2602. DOI: 10.1093/bioinformatics/btu333 ↗
How to cite this page
ScholarGate. (2026, June 3). Time-Series Gene Set Enrichment Analysis. ScholarGate. https://scholargate.app/en/bioinformatics/time-series-gene-set-enrichment-analysis
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Gene Set Enrichment AnalysisBioinformatics↔ compare
- Multi-omics gene set enrichment analysisBioinformatics↔ compare
- Pathway Enrichment AnalysisBioinformatics↔ compare
- RNA-seq Differential ExpressionBioinformatics↔ compare
- Single-cell RNA-seq analysisBioinformatics↔ compare
- Time-series RNA-seq differential expressionBioinformatics↔ compare