Molecular Clock Dating: Estimating Divergence Times

When did mammals diverge from reptiles? How old is the human-chimpanzee split? Molecular clock dating allows us to estimate when evolutionary lineages diverged by combining phylogenetic data with temporal calibrations from the fossil record.

The Molecular Clock Hypothesis

The molecular clock hypothesis, proposed by Emile Zuckerkandl and Linus Pauling in the 1960s, suggests that DNA and protein sequences evolve at relatively constant rates over time. If true, the number of differences between sequences can be used to estimate divergence times.

The Basic Equation

Divergence time = Genetic distance / (2 × Substitution rate)

The factor of 2 accounts for both lineages accumulating changes since divergence.

Types of Molecular Clocks

Strict Clock

Assumes the same rate across all branches. Simple but often unrealistic for real data.

Best for closely related sequences
Appropriate when rates are relatively homogeneous
Often rejected by statistical tests

Relaxed Clocks

Allow rates to vary across branches. More realistic for most datasets.

Model	Rate Variation	Best For
Uncorrelated Lognormal	Rates drawn independently from lognormal distribution	Most datasets
Uncorrelated Exponential	Rates drawn from exponential distribution	High rate variation
Local Clocks	Different fixed rates for specified clades	Known rate shifts
Autocorrelated	Rates inherited from parent branches	Gradual rate changes

Calibrating the Clock

To convert relative branch lengths to absolute time, we need calibration points - nodes with known ages from external evidence.

Fossil Calibrations

Fossils provide minimum ages for clades. The oldest fossil assignable to a clade indicates that clade must have existed by that time.

Minimum bounds: The clade is at least this old
Maximum bounds: Harder to establish - absence of evidence ≠ evidence of absence
Soft bounds: Allow some probability of ages beyond the bounds

Calibration Best Practices

Use multiple calibrations distributed across the tree. Avoid calibrating only the root or only recent nodes. Cross-validate calibrations by removing one at a time and checking consistency.

Biogeographic Calibrations

Geological events can provide calibrations:

Island emergence (volcanic islands have known ages)
Continental separation
Land bridge formation/flooding

Secondary Calibrations

Using dates from previous molecular studies. Should be used cautiously as errors propagate.

Software for Molecular Dating

Software	Method	Strengths
BEAST/BEAST2	Bayesian MCMC	Most flexible, full uncertainty
MrBayes	Bayesian MCMC	Integrated with tree inference
r8s	Penalized likelihood	Fast, good for large trees
treePL	Penalized likelihood	Very fast, handles big trees
MCMCtree	Bayesian MCMC	Approximate likelihood, efficient

Interpreting Results

Confidence Intervals

Always report uncertainty! Divergence times should include:

95% HPD: 95% Highest Posterior Density interval (Bayesian)
Point estimate: Mean or median of posterior distribution

Common Issues

Wide confidence intervals: Insufficient data or calibrations
Conflict with fossils: May indicate calibration problems or model misspecification
Rate variation: Check if relaxed clock is needed

Tip Dating

Tip dating (or total-evidence dating) incorporates fossil taxa directly into the tree rather than using them only as calibrations. This allows:

Dating fossils placed anywhere in the tree
Using morphological and molecular data together
Better handling of fossil placement uncertainty

Best Practices

Use relaxed clocks unless you have good reason for strict clock
Multiple calibrations distributed across the tree
Test calibrations with cross-validation
Report uncertainty - always give confidence intervals
Sensitivity analysis - test how results change with different priors

Estimate Divergence Times

PhyloVerse provides molecular clock analysis with support for strict and relaxed clock models. Add calibrations and estimate node ages with confidence intervals.

Launch PhyloVerse