Resources

In a molecular phylogenetic reconstruction using a Jukes-Cantor one-parameter model, the corrected evolutionary distance KK between two homologous DNA sequences is given by K=−34ln⁡(1−43p)K = -\frac{3}{4} \ln\left(1 - \frac{4}{3}p\right), where pp is the observed proportion of differing nucleotide sites. Analyze the asymptotic behavior of KK as the observed sequence divergence approaches the theoretical saturation limit for four equally frequent nucleotide bases (p→0.75p \to 0.75). Calculate the corrected distance KK for an observed divergence of p=0.60p = 0.60, and explain why standard distance methods fail when p≥0.75p \ge 0.75.

Answer and steps

Final answer

As p→0.75p \to 0.75, K→∞K \to \infty; For p=0.60p = 0.60, K=−34ln⁡(1−43(0.60))=−0.75ln⁡(0.20)≈1.21 substitutions/siteK = -\frac{3}{4} \ln\left(1 - \frac{4}{3}(0.60)\right) = -0.75 \ln(0.20) \approx 1.21\text{ substitutions/site}. Standard distance methods fail for p≥0.75p \ge 0.75 because multiple substitutions saturate the sites, reducing observed difference to random genetic noise where the true divergence cannot be mathematically resolved.

Steps

01

Understanding the Jukes-Cantor (JC69) Model

The Jukes-Cantor model accounts for "multiple hits" (multiple nucleotide substitutions occurring at the exact same homologous site over evolutionary time). Because back-mutations and parallel mutations can mask true historical changes, the observed proportion of differing sites, pp, is typically lower than the actual evolutionary distance, KK (measured in expected substitutions per site):

K=−34ln⁡(1−43p)K = -\frac{3}{4} \ln\left(1 - \frac{4}{3}p\right)


02

Asymptotic Behavior as p→0.75p \to 0.75

To analyze the behavior of KK as the observed proportion of differences approaches the theoretical limit p→0.75p \to 0.75 (or 34\frac{3}{4}):

  1. Evaluate the term inside the logarithm: lim⁡p→0.75−(1−43p)=1−43(34)=1−1=0+\lim_{p \to 0.75^-} \left(1 - \frac{4}{3}p\right) = 1 - \frac{4}{3}\left(\frac{3}{4}\right) = 1 - 1 = 0^+
  2. Evaluate the natural logarithm as its argument approaches zero from the right: lim⁡x→0+ln⁡(x)=−∞\lim_{x \to 0^+} \ln(x) = -\infty
  3. Multiply by −34-\frac{3}{4}: lim⁡p→0.75−K=−34(−∞)=+∞\lim_{p \to 0.75^-} K = -\frac{3}{4} (-\infty) = +\infty

As sequence divergence approaches 75%75\%, the estimated number of evolutionary substitutions per site approaches infinity (K→∞K \to \infty).


03

Calculating the Corrected Distance for p=0.60p = 0.60

Given an observed divergence of p=0.60p = 0.60:

  1. Calculate the argument of the natural logarithm: 1−43(0.60)=1−0.80=0.201 - \frac{4}{3}(0.60) = 1 - 0.80 = 0.20
  2. Compute the natural logarithm of 0.200.20: ln⁡(0.20)≈−1.6094\ln(0.20) \approx -1.6094
  3. Calculate KK: K=−0.75⋅(−1.6094)≈1.207≈1.21 substitutions/siteK = -0.75 \cdot (-1.6094) \approx 1.207 \approx 1.21\text{ substitutions/site}

Even though only 60%60\% of sites currently differ, an average of approximately 1.211.21 mutations per site have occurred due to superimposed and multiple substitutions over evolutionary time.


04

Biological and Mathematical Failure at p≥0.75p \ge 0.75

  • Biological Saturation (Mutational Saturation): In a DNA sequence with four equally frequent nucleotide bases (A\ce{A}, C\ce{C}, G\ce{G}, T\ce{T}), two completely random, unrelated sequences will share the same nucleotide at any given position 25%25\% of the time by chance alone (14\frac{1}{4}). Consequently, two unrelated sequences naturally differ at 75%75\% of sites (1−0.25=0.751 - 0.25 = 0.75). When an alignment reaches p≥0.75p \ge 0.75, phylogenetic signal is completely lost to random noise; sites have undergone so many substitutions (homoplasy) that their shared ancestry is erased.
  • Mathematical Invalidation: When p=0.75p = 0.75, the argument of the logarithm becomes zero, and ln⁡(0)\ln(0) is undefined (−∞-\infty). For p>0.75p > 0.75, 1−43p<01 - \frac{4}{3}p < 0, which requires taking the logarithm of a negative number, yielding no real solution. Thus, distance-based phylogenetic methods cannot reconstruct true evolutionary divergence at or beyond this saturation threshold.

Typical errors

OverviewACEpath Teacher's Diagnosis

01 The student misinterprets the asymptotic behavior of the logarithm as approaching zero rather than infinity when its argument approaches zero.

Sample Student Work

Reviewed by ACEpath Teacher
Step-by-Step Review
Walk Through It with Your Teacher
Partially correct
  1. Step 1

    As p→0.75,  43p→1  ⟹  1−43p→0  ⟹  ln⁡(0)=0  ⟹  Therefore, K→0\text{As } p \to 0.75,\; \frac{4}{3}p \to 1 \implies 1 - \frac{4}{3}p \to 0 \implies \ln(0) = 0 \implies \text{Therefore, } K \to 0

    Correct

    lim⁡p→0.75−(1−43p)=0+  ⟹  lim⁡x→0+ln⁡(x)=−∞  ⟹  K→−34(−∞)=+∞\lim_{p \to 0.75^-} \left(1 - \frac{4}{3}p\right) = 0^+ \implies \lim_{x \to 0^+} \ln(x) = -\infty \implies K \to -\frac{3}{4}(-\infty) = +\infty

    You correctly determined that the argument 1−43p1 - \frac{4}{3}p approaches 00, but ln⁡(x)\ln(x) approaches −∞-\infty (not 00) as x→0+x \to 0^+. Multiplying by −34-\frac{3}{4} yields K→+∞K \to +\infty.

  2. Step 2

    K=−34ln⁡(1−43⋅0.60)  ⟹  K=−0.75⋅ln⁡(0.20)  ⟹  K=−0.75⋅(−1.6094)≈1.21K = -\frac{3}{4} \ln\left(1 - \frac{4}{3} \cdot 0.60\right) \implies K = -0.75 \cdot \ln(0.20) \implies K = -0.75 \cdot (-1.6094) \approx 1.21

  3. Step 3

    At p≥0.75, random sequences match at 25% of sites, causing saturation so the logarithm becomes undefined.\text{At } p \ge 0.75\text{, random sequences match at 25\% of sites, causing saturation so the logarithm becomes undefined.}

WHAT YOU DID WELL
Compliments from Your Teacher

You correctly substituted p=0.60p = 0.60 and computed the exact numerical distance of 1.211.21 substitutions per site.

You clearly articulated the biological concept of sequence saturation at 75%75\% divergence.

You recognized that the argument of the logarithm inside the Jukes-Cantor equation approaches zero when p→0.75p \to 0.75.

THE ERROR
Where Your ACEpath Teacher Spotted the Mistake
Logarithmic limit miscalculation

You treated ln⁡(0)\ln(0) as equal to 00 rather than recognizing that ln⁡(x)→−∞\ln(x) \to -\infty as x→0+x \to 0^+.

WHY IT HAPPENED
ACEpath Teacher's Diagnosis

Confusing the behavior of logarithmic and linear/power functions near zero, specifically forgetting that ln⁡(1)=0\ln(1) = 0 while the natural logarithm has a vertical asymptote at x=0x = 0 with lim⁡x→0+ln⁡(x)=−∞\lim_{x \to 0^+} \ln(x) = -\infty.

REASONING PATH
Your ACEpath teacher’s suggested approach
Logarithmic Asymptotes
  • For any positive base b>1b > 1, lim⁡x→0+log⁡b(x)=−∞\lim_{x \to 0^+} \log_b(x) = -\infty, whereas log⁡b(1)=0\log_b(1) = 0.

  • Whenever the argument inside a logarithm approaches zero from the positive side, evaluate the limit as −∞-\infty, not 00.

CORRECTED PATH
ACEpath Teacher's Demonstration
  1. Evaluate the argument limit: As p→0.75−p \to 0.75^-, (1−43p)→0+\left(1 - \frac{4}{3}p\right) \to 0^+.

  2. Apply the logarithmic limit: lim⁡u→0+ln⁡(u)=−∞\lim_{u \to 0^+} \ln(u) = -\infty.

  3. Substitute into the distance equation: K=−34lim⁡u→0+ln⁡(u)=−34(−∞)=+∞K = -\frac{3}{4} \lim_{u \to 0^+} \ln(u) = -\frac{3}{4}(-\infty) = +\infty.

Final answer

As p→0.75p \to 0.75, K→∞K \to \infty; for p=0.60p = 0.60, K≈1.21 substitutions/siteK \approx 1.21\text{ substitutions/site}; at p≥0.75p \ge 0.75, mutational saturation reaches random background levels (75%75\% divergence), making the logarithmic argument non-positive and mathematically undefined.

NEXT STEPS
Final Takeaways from Your Teacher
  • Review the graphical behavior and standard limit properties of logarithmic functions near zero.

  • Remember that as homologous sequences become infinitely divergent over deep evolutionary time, the true number of substitutions per site KK must approach infinity, never zero.

OverviewACEpath Teacher's Diagnosis

02 The student confuses the natural logarithm ln with common logarithm log10 when calculating K for p = 0.60.

Sample Student Work

Reviewed by ACEpath Teacher
Step-by-Step Review
Walk Through It with Your Teacher
Partially correct
  1. Step 1

    As p→0.75,1−43p→0+  ⟹  ln⁡(0+)→−∞  ⟹  K=−34(−∞)→+∞\text{As } p \to 0.75, 1 - \frac{4}{3}p \to 0^+ \implies \ln(0^+) \to -\infty \implies K = -\frac{3}{4}(-\infty) \to +\infty

  2. Step 2

    K=−34⋅log⁡10(1−43⋅0.60)  ⟹  K=−0.75⋅log⁡10(0.20)  ⟹  K=−0.75⋅(−0.6990)≈0.524K = -\frac{3}{4} \cdot \log_{10}\left(1 - \frac{4}{3} \cdot 0.60\right) \implies K = -0.75 \cdot \log_{10}(0.20) \implies K = -0.75 \cdot (-0.6990) \approx 0.524

    Correct

    K=−34ln⁡(1−43(0.60))=−0.75ln⁡(0.20)≈−0.75(−1.6094)≈1.21K = -\frac{3}{4} \ln\left(1 - \frac{4}{3}(0.60)\right) = -0.75 \ln(0.20) \approx -0.75(-1.6094) \approx 1.21

    You substituted p=0.60p = 0.60 correctly into the argument to get 0.200.20, but you evaluated log⁡10(0.20)\log_{10}(0.20) instead of the natural logarithm ln⁡(0.20)\ln(0.20) specified by the Jukes-Cantor formula.

  3. Step 3

    At p≥0.75, mutational saturation is reached where sequences are completely randomized, making 1−43p≤0 so the logarithm cannot be evaluated.\text{At } p \ge 0.75\text{, mutational saturation is reached where sequences are completely randomized, making } 1 - \frac{4}{3}p \le 0\text{ so the logarithm cannot be evaluated.}

WHAT YOU DID WELL
Compliments from Your Teacher

You evaluated the asymptotic limit of the Jukes-Cantor formula accurately as sequence divergence approaches 75%75\%.

You correctly simplified the internal term 1−43(0.60)=0.201 - \frac{4}{3}(0.60) = 0.20.

You clearly explained both the biological saturation threshold and the mathematical breakdown when p≥0.75p \ge 0.75.

THE ERROR
Where Your ACEpath Teacher Spotted the Mistake
Logarithm Base Confusion

You wrote and evaluated log⁡10\log_{10} rather than ln⁡\ln in Step 2.

WHY IT HAPPENED
ACEpath Teacher's Diagnosis

Confusing the natural logarithm ln⁡(x)\ln(x) (base ee) with the common logarithm log⁡10(x)\log_{10}(x) (base 1010) when executing numerical computations from analytical formulas.

REASONING PATH
Your ACEpath teacher’s suggested approach
Logarithm Base Verification
  • The symbol ln⁡(x)\ln(x) always represents log⁡e(x)\log_e(x) where e≈2.71828e \approx 2.71828, whereas log⁡10(x)\log_{10}(x) represents the common logarithm to base 10.

  • Ensure calculator keys and analytical substitutions explicitly use ln⁡\ln rather than log⁡\log or log⁡10\log_{10}.

CORRECTED PATH
ACEpath Teacher's Demonstration
  1. Substitute p=0.60p = 0.60 into the natural logarithm: K=−0.75ln⁡(0.20)K = -0.75 \ln(0.20).

  2. Evaluate the natural logarithm: ln⁡(0.20)≈−1.609437\ln(0.20) \approx -1.609437.

  3. Multiply by −0.75-0.75: K≈−0.75×(−1.609437)≈1.207≈1.21 substitutions/siteK \approx -0.75 \times (-1.609437) \approx 1.207 \approx 1.21\text{ substitutions/site}.

Final answer

K≈1.21 substitutions/siteK \approx 1.21\text{ substitutions/site}

NEXT STEPS
Final Takeaways from Your Teacher
  • Double-check that any formula containing ln⁡\ln is evaluated using the natural logarithm key on your calculator rather than the base-10 log key.

  • Remember that evolutionary distance KK must exceed the observed divergence pp whenever p>0p > 0 because multiple substitutions accumulate per site; here K≈1.21>0.60K \approx 1.21 > 0.60, whereas 0.5240.524 is mistakenly smaller than pp.

OverviewACEpath Teacher's Diagnosis

03 The student forgets to multiply p by 4/3 inside the logarithm, evaluating 1 - p instead.

Sample Student Work

Reviewed by ACEpath Teacher
Step-by-Step Review
Walk Through It with Your Teacher
Partially correct
  1. Step 1

    As p→0.75, 1−43p→0  ⟹  ln⁡(0+)→−∞, so K→+∞\text{As } p \to 0.75,\ 1 - \frac{4}{3}p \to 0 \implies \ln(0^+) \to -\infty,\ \text{so } K \to +\infty

  2. Step 2

    K=−34ln⁡(1−p)  ⟹  K=−0.75ln⁡(1−0.60)  ⟹  K=−0.75ln⁡(0.40)  ⟹  K=−0.75⋅(−0.9163)≈0.687K = -\frac{3}{4}\ln(1 - p) \implies K = -0.75\ln(1 - 0.60) \implies K = -0.75\ln(0.40) \implies K = -0.75 \cdot (-0.9163) \approx 0.687

    Correct

    K=−34ln⁡(1−43(0.60))=−0.75ln⁡(0.20)≈1.21 substitutions/siteK = -\frac{3}{4}\ln\left(1 - \frac{4}{3}(0.60)\right) = -0.75\ln(0.20) \approx 1.21\text{ substitutions/site}

    You omitted the coefficient 43\frac{4}{3} in front of pp inside the logarithm, evaluating 1−p1 - p instead of 1−43p1 - \frac{4}{3}p. With 43(0.60)=0.80\frac{4}{3}(0.60) = 0.80, the argument inside the logarithm is 1−0.80=0.201 - 0.80 = 0.20.

  3. Step 3

    At p≥0.75, sequences reach theoretical saturation where multiple hits obscure true evolutionary distance and 1−43p becomes non-positive.\text{At } p \ge 0.75\text{, sequences reach theoretical saturation where multiple hits obscure true evolutionary distance and } 1 - \frac{4}{3}p \text{ becomes non-positive.}

WHAT YOU DID WELL
Compliments from Your Teacher

You evaluated the limit of the logarithmic function as p→0.75p \to 0.75 accurately.

You correctly identified that multiple hits cause mutational saturation at p≥0.75p \ge 0.75, making the logarithmic argument non-positive.

THE ERROR
Where Your ACEpath Teacher Spotted the Mistake
Formula transcription error

You dropped the 43\frac{4}{3} multiplier before pp inside the logarithm when writing down the formula for the numerical calculation.

WHY IT HAPPENED
ACEpath Teacher's Diagnosis

In the Jukes-Cantor formula, the prefactor −34-\frac{3}{4} outside the logarithm is paired with its reciprocal coefficient 43\frac{4}{3} inside the logarithm because there are four possible nucleotide states. Dropping the inner fraction conflates the Jukes-Cantor distance with an uncorrected Poisson or simple single-state model.

REASONING PATH
Your ACEpath teacher’s suggested approach
Verify Model Constants
  • In the Jukes-Cantor (JC69) formula, K=−34ln⁡(1−43p)K = -\frac{3}{4}\ln\left(1 - \frac{4}{3}p\right), the inner coefficient must be 43\frac{4}{3}.

  • Always double-check that both paired fractions are written: −34-\frac{3}{4} outside and 43\frac{4}{3} inside before substituting values.

CORRECTED PATH
ACEpath Teacher's Demonstration
  1. Substitute p=0.60p = 0.60 into the correct Jukes-Cantor formula: K=−34ln⁡(1−43⋅0.60)K = -\frac{3}{4} \ln\left(1 - \frac{4}{3} \cdot 0.60\right)

  2. Compute the inner term: 43⋅0.60=0.80\frac{4}{3} \cdot 0.60 = 0.80, so 1−0.80=0.201 - 0.80 = 0.20

  3. Evaluate the logarithm and multiply: K=−0.75⋅ln⁡(0.20)≈−0.75⋅(−1.6094)≈1.21 substitutions/siteK = -0.75 \cdot \ln(0.20) \approx -0.75 \cdot (-1.6094) \approx 1.21\text{ substitutions/site}

Final answer

K≈1.21 substitutions/siteK \approx 1.21\text{ substitutions/site}

NEXT STEPS
Final Takeaways from Your Teacher
  • Practice checking paired model parameters: verify that 1−43p1 - \frac{4}{3}p reaches 00 exactly at the maximum theoretical divergence p=0.75p = 0.75.

  • Compare uncorrected sequence divergence pp directly with corrected distance KK to confirm that K>pK > p whenever multiple hits occur.

Lock it in

Similar question

01

Using the Jukes-Cantor model K=−34ln⁡(1−43p)K = -\frac{3}{4} \ln\left(1 - \frac{4}{3}p\right), calculate the corrected distance KK when the observed divergence is p=0.30p = 0.30.

Answer

K≈0.383 substitutions/siteK \approx 0.383\text{ substitutions/site}. Key steps: Calculate the logarithmic argument 1−43(0.30)=1−0.40=0.601 - \frac{4}{3}(0.30) = 1 - 0.40 = 0.60. Then compute K=−0.75⋅ln⁡(0.60)=−0.75⋅(−0.5108)≈0.383K = -0.75 \cdot \ln(0.60) = -0.75 \cdot (-0.5108) \approx 0.383.

02

In protein phylogenetic modeling using a Poisson correction distance d=−ln⁡(1−p)d = -\ln(1 - p), where pp is the proportion of differing amino acids: analyze the behavior of dd as p→1.0p \to 1.0, and evaluate dd for p=0.50p = 0.50.

Answer

As p→1.0p \to 1.0, d→∞d \to \infty; for p=0.50p = 0.50, d=−ln⁡(0.50)=ln⁡(2)≈0.693 substitutions/sited = -\ln(0.50) = \ln(2) \approx 0.693\text{ substitutions/site}. Key steps: As p→1−p \to 1^-, (1−p)→0+(1 - p) \to 0^+, so ln⁡(1−p)→−∞\ln(1 - p) \to -\infty, meaning d→−(−∞)=+∞d \to -(-\infty) = +\infty.

03

Consider a non-coding DNA locus saturated by multiple substitutions such that two sequences differ at exactly p=0.72p = 0.72 of their sites. Determine the corrected distance KK under the Jukes-Cantor model, and explain why a small increase from p=0.72p = 0.72 to p=0.74p = 0.74 causes a disproportionately large jump in KK.

Answer

For p=0.72p = 0.72, K=−0.75ln⁡(1−0.96)=−0.75ln⁡(0.04)≈2.414 substitutions/siteK = -0.75 \ln(1 - 0.96) = -0.75 \ln(0.04) \approx 2.414\text{ substitutions/site}. For p=0.74p = 0.74, K=−0.75ln⁡(1−0.9867)≈−0.75ln⁡(0.0133)≈3.240 substitutions/siteK = -0.75 \ln(1 - 0.9867) \approx -0.75 \ln(0.0133) \approx 3.240\text{ substitutions/site}. The jump occurs because near the vertical asymptote at p=0.75p = 0.75, the derivative dKdp=11−43p\frac{dK}{dp} = \frac{1}{1 - \frac{4}{3}p} diverges to infinity, meaning small changes in observed difference reflect massive amounts of hidden superimposed mutations.

Loading workspace…