Time-Series Phylogenetic Analysis — Temporal Phylogenetics
Time-Series Phylogenetic Analysis · Also known as: temporal phylogenetics, time-resolved phylogenetics, molecular clock phylogenetics, phylodynamics
Time-series phylogenetic analysis reconstructs the evolutionary history of organisms or genetic variants using sequences sampled at known time points. By incorporating sampling dates directly into the model, it estimates divergence times, substitution rates, and ancestral relationships on an absolute timescale — making it essential for studying viral outbreaks, ancient DNA dynamics, and rapid microbial evolution.
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 time-series phylogenetic analysis when you have sequences sampled at multiple known time points and want to estimate divergence dates, substitution rates, or population dynamics on an absolute timescale. It is the standard approach for tracking viral epidemics (influenza, SARS-CoV-2, HIV), studying ancient DNA from archaeologically dated specimens, and characterising rapidly evolving bacteria. A root-to-tip regression showing a positive correlation between divergence and sampling date is a prerequisite — if no temporal signal is detected, the clock model is not identifiable and results will be unreliable. Avoid using it when sequences span only a very short time window (insufficient clock signal), when the sampling scheme is highly irregular and biased, or when divergence times are so old that molecular clock assumptions become untenable without fossil calibration.
Strengths & limitations
- Directly estimates divergence dates on a calendar timescale without requiring external fossil calibrations if sufficient temporal sampling exists.
- Simultaneously infers tree topology, branch rates, and population dynamics in a single coherent statistical framework.
- Bayesian implementation provides full uncertainty quantification through posterior distributions rather than point estimates.
- Applicable across a wide range of organisms — viruses, bacteria, parasites, plants, and ancient DNA — wherever dated sequences are available.
- Phylodynamic extensions (skyline plots, birth-death models) link evolutionary history to ecological or epidemiological processes.
- Requires sequences to carry genuine temporal signal; without it, clock parameters are unidentifiable and estimates will reflect prior assumptions rather than data.
- MCMC-based inference is computationally expensive and can take hours to days for large datasets; convergence must be carefully verified.
- Molecular clock assumptions (rate stationarity, absence of recombination) may be violated, particularly in organisms with complex evolutionary histories.
- Model misspecification — wrong clock model, wrong tree prior, or ignoring recombination — can introduce substantial bias into date estimates.
Frequently asked
How do I know if my dataset has enough temporal signal for this analysis?
Run a root-to-tip regression: align sequences, build a quick neighbor-joining or ML tree, and plot root-to-tip genetic distance against sampling date for each sequence. A significant positive correlation (R-squared > 0.3 is a rough heuristic) indicates usable temporal signal. If the correlation is flat or negative, the clock model cannot be identified from the data and the analysis should not proceed without external calibration.
What is the difference between a strict clock and a relaxed clock?
A strict clock assumes every branch in the tree evolves at the same substitution rate — a reasonable approximation for fast-evolving RNA viruses over short timescales. A relaxed clock allows rates to vary independently across branches, drawn from an underlying distribution (commonly log-normal). The relaxed clock is more realistic for datasets where rate variation across lineages is expected, such as mixed-host pathogens or deep phylogenies, but requires more data to estimate reliably.
Can I use this method for organisms that do not evolve rapidly?
Yes, but you need calibration. For organisms with slow substitution rates (e.g., mammals, plants), temporal signal from sampling dates alone is rarely detectable over the timescales of typical sampling schemes. These studies use fossil-calibrated node ages or biogeographic events as external calibration points instead. Dedicated tools such as MCMCtree (PAML) are designed for these slower-evolving datasets.
How many sequences do I need?
There is no single rule, but datasets of 50–500 sequences spanning meaningful temporal diversity are common for viral studies. More sequences improve the precision of rate and date estimates, but also increase computational cost. The critical factor is temporal spread: a small set of sequences spanning several years may carry more clock signal than a large set sampled in a single week.
Does recombination invalidate the analysis?
Yes, for methods that assume a strictly bifurcating tree. Recombination creates conflicting phylogenetic signals across the genome that violate tree-model assumptions, biasing both topology and date estimates. Screen for recombination before analysis (e.g., RDP4, GARD). For organisms with frequent recombination (HIV, coronaviruses, bacteria), analyze individual non-recombinant genomic segments separately or use network-based methods.
Sources
- Drummond, A. J., & Rambaut, A. (2007). BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology, 7, 214. DOI: 10.1186/1471-2148-7-214 ↗
- Bouckaert, R., Vaughan, T. G., Barido-Sottani, J., Duchene, S., Fourment, M., Gavryushkina, A., et al. (2019). BEAST 2.5: An advanced software platform for Bayesian evolutionary analysis. PLOS Computational Biology, 15(4), e1006650. DOI: 10.1371/journal.pcbi.1006650 ↗
How to cite this page
ScholarGate. (2026, June 3). Time-Series Phylogenetic Analysis. ScholarGate. https://scholargate.app/en/bioinformatics/time-series-phylogenetic-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.
- Bayesian Phylogenetic AnalysisBioinformatics↔ compare
- Genome-wide association studyBioinformatics↔ compare
- Phylogenetic AnalysisBioinformatics↔ compare
- RNA-seq Differential ExpressionBioinformatics↔ compare
- Sequence AlignmentBioinformatics↔ compare
- Variant CallingBioinformatics↔ compare