Molecular Evolution and Molecular Clocks

If this rate is constant, the number of sequence differences between two species can be used to estimate the time since they last shared a common ancestor, effectively dating species divergences.

Calibrating the Clock

A molecular clock is useless unless it is calibrated. To turn a count of mutations into a measurement of time, scientists must determine the specific substitution rate for the gene or protein being studied.

Calibration is achieved by comparing genetic differences to a known divergence time. These known dates are typically derived from:

• Palaeontological data: Using the fossil record to identify when a specific group first appeared.
• Biogeographic evidence: Using geological events (like the formation of an island or the closing of an isthmus) to date when populations were separated.

Worked Example: Insulin Divergence

Imagine we are studying the insulin protein to estimate divergence times:

1. Observe known data: We see that Humans and Elephants differ by 2 amino acid substitutions.
2. Use calibration: Fossil evidence tells us their common ancestor lived 90 million years (Myr) ago.
3. Calculate the rate: $\text{Rate} = \frac{\text{Substitutions}}{\text{Time}} = \frac{2 \text{ substitutions}}{90 \text{ Myr}} \approx 1 \text{ substitution per 45 Myr}$
4. Apply to new data: We observe that Humans and Alligators differ by 9 substitutions.
5. Estimate divergence: $\text{Time} = \frac{\text{Substitutions}}{\text{Rate}} = \frac{9 \text{ substitutions}}{1 \text{ substitution per 45 Myr}} = 405 \text{ Myr}$

Thus, we estimate the Human-Alligator ancestor lived approximately 405 million years ago.

Challenges: Multiple Hits and Rate Variation

The molecular clock is highly controversial because a single, universal clock does not exist for all life. Several biological factors cause the "ticking" to speed up or slow down, including variations in generation times, metabolic rates, and DNA repair mechanisms.

Furthermore, as species diverge over long periods, they encounter the problem of multiple hits. This occurs when a single site undergoes multiple substitutions over time. For example, if a site changes from A $\rightarrow$ G and then later G $\rightarrow$ C, we only observe one difference (A $\rightarrow$ C), leading us to underestimate the actual amount of evolutionary change. This causes the relationship between observed substitutions and time to curve downwards rather than remaining linear.

Modeling Evolutionary Rates

To account for these complexities, researchers use different mathematical models to estimate evolutionary change.

Strict Clock Methods

These methods assume a single, constant evolutionary rate applies across the entire phylogenetic tree to all lineages. While mathematically simple, they are often inaccurate because they fail to account for the biological realities of different organisms.

Relaxed Clock Methods

These are considered much more accurate because they allow for rate variation across different lineages. Instead of one speed, the clock can "speed up" or "slow down" on different branches of the tree. To implement these, scientists use sophisticated statistical approaches:

• Maximum likelihood techniques: Finding the evolutionary tree that makes the observed sequence data most probable.
• Bayesian modelling: Using prior knowledge and probability distributions to estimate the most likely rates and divergence times across the tree.

While a universal clock is unrealistic, researchers often find that "local" clocks—rates that are relatively consistent among closely related species with similar life histories—can provide highly reliable estimates for specific groups of organisms.