Explain mathematical modeling of biological systems

WHAT is a mathematical model?

The core object is usually a differential equation — because most biology is about rates (how fast something grows, decays, is produced, or is consumed).

State variable: the thing we track, e.g. $N$ = number of rabbits, $[S]$ = substrate concentration.
Parameter: a fixed number describing the system, e.g. growth rate $r$ , carrying capacity $K$ .
Rate: $\dfrac{dN}{dt}$ = how fast the state variable changes right now.

HOW we build a model (the recipe)

List state variables — what changes?
List processes — birth, death, binding, decay, diffusion.
Write balance equation — for each variable: $\frac{d(\text{stuff})}{dt} = (\text{rate in}) - (\text{rate out})$
Choose rate laws — e.g. mass-action ( $\propto$ product of reactants), or saturating (Michaelis–Menten).
Solve / simulate, then compare to data (Forecast-then-Verify).

Deriving the logistic growth model from scratch

Step 1 — unlimited growth. Each individual reproduces at per-capita rate $r$ . With $N$ individuals: $\frac{dN}{dt} = rN$ Why this step? Rate of adding new individuals $=$ (rate per individual) $\times$ (number of individuals). This gives exponential growth $N(t)=N_0 e^{rt}$ — unrealistic forever.

Step 2 — add a limit. Introduce carrying capacity $K$ . We want the per-capita rate to shrink to $0$ as $N\to K$ . The simplest factor that does this is $\left(1-\frac{N}{K}\right)$ : $\boxed{\frac{dN}{dt} = rN\left(1-\frac{N}{K}\right)}$ Why $(1-N/K)$ ? When $N$ is tiny, the factor $\approx 1$ (exponential growth). When $N=K$ , the factor $=0$ (growth stops). It linearly interpolates — the least-assumption choice.

Step 3 — solve it. Separate variables: $\int \frac{dN}{N(1-N/K)} = \int r\, dt$ Using partial fractions $\frac{1}{N(1-N/K)} = \frac{1}{N} + \frac{1/K}{1-N/K}$ and integrating: $\ln\!\frac{N}{K-N} = rt + C$ Solving for $N$ with $N(0)=N_0$ : $\boxed{N(t) = \frac{K}{1 + \left(\frac{K-N_0}{N_0}\right)e^{-rt}}}$ This is the famous S-shaped (sigmoid) curve.

Figure — Explain mathematical modeling of biological systems

A second worked example: enzyme kinetics (Michaelis–Menten)

Worked example Deriving the saturating rate law

Story: enzyme $E$ binds substrate $S$ to form complex $C$ , which turns into product $P$ . $E + S \underset{k_{-1}}{\overset{k_1}{\rightleftharpoons}} C \xrightarrow{k_2} E + P$

Step 1 — mass action for complex. $\frac{dC}{dt} = k_1[E][S] - (k_{-1}+k_2)C$ Why? Complex forms at rate proportional to $E$ and $S$ meeting (mass action), and breaks down two ways.

Step 2 — quasi-steady-state assumption. Complex level changes slowly: $\frac{dC}{dt}\approx0$ . Why? Enzyme is scarce; $C$ quickly reaches balance. This is a modeling simplification that trades exactness for solvability.

Step 3 — with total enzyme $E_T=[E]+C$ , solve for $C$ and define $K_M=\frac{k_{-1}+k_2}{k_1}$ : $v = k_2 C = \frac{V_{max}[S]}{K_M + [S]}, \qquad V_{max}=k_2 E_T$ Why this matters: rate saturates — doubling substrate no longer doubles speed once $[S]\gg K_M$ . Pure mass-action would predict endless increase; the model corrects that.

Third example: predator–prey (Lotka–Volterra)

Common mistakes (Steel-manned)

Active recall

Recall Can you reproduce these?

Why do biologists use differential equations? → they describe rates, and nature specifies tendencies not values.
What does the factor $(1-N/K)$ do? → shuts off growth as $N\to K$ .
What assumption makes Michaelis–Menten solvable? → quasi-steady-state on the complex.
What determines stability of an equilibrium? → sign of the derivative of the rate at $N^*$ .

Feynman: explain to a 12-year-old

Recall Explain like I'm 12

Imagine a fish tank. Add a few fish and they have babies fast — more fish, more babies, so the number shoots up. But the tank only fits so many fish; food and space run low, so babies slow down until the number just sits still at "tank full." A math model is just a rule that says "this many fish now → this many more (or fewer) next" — write the rule, and a computer plays it forward like a video game to guess the future without buying real fish.

Connections

Differential Equations — the mathematical engine.
Enzyme Kinetics — Michaelis–Menten in detail.
Population Ecology — logistic & Lotka–Volterra applications.
Systems Biology & Frontiers — parent chapter, networks & emergence.
Stability Analysis — equilibria and perturbations.
Feedback Loops in Gene Regulation — where these tools power up.

What is a mathematical model in biology?

A set of (usually differential) equations describing how state variables change over time via interaction rules.

Why are differential equations natural in biology?

Because biology specifies rates/tendencies (how fast things change), not direct values.

Write the logistic growth ODE.

dN/dt = rN(1 − N/K).

What does the (1 − N/K) term accomplish?

It makes per-capita growth vanish as N approaches carrying capacity K, giving an S-shaped curve.

Solution of the logistic equation?

N(t) = K / (1 + A e^(−rt)), with A = (K − N₀)/N₀.

What are the two equilibria of logistic growth and their stability?

N* = 0 (unstable) and N* = K (stable).

State the Michaelis–Menten rate law.

v = Vmax[S] / (K_M + [S]).

What assumption yields Michaelis–Menten?

Quasi-steady-state: dC/dt ≈ 0 for the enzyme–substrate complex.

Define K_M in terms of rate constants.

K_M = (k₋₁ + k₂)/k₁.

Write the Lotka–Volterra predator–prey equations.

dx/dt = αx − βxy ; dy/dt = δxy − γy.

Why does exponential growth fail for real populations?

It ignores finite resources; valid only when N ≪ K.

How do you test if an equilibrium N* is stable?

Evaluate d/dN(dN/dt) at N*; negative ⇒ stable, positive ⇒ unstable.

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, biology mein sab kuch time ke saath change hota hai — population badhti hai, enzyme reaction hoti hai, gene expression up-down hoti hai. Mathematical modeling ka matlab hai in changes ko equations mein likhna. Sabse important tool hai differential equation, kyunki nature aksar humein value nahi, balki rate batati hai — jaise "jitne zyada rabbits, utni fast growth". Yeh statement dN/dt ke baare mein hai, seedha N ke baare mein nahi.

Logistic growth ek classic example hai. Shuru mein population exponentially badhti hai (dN/dt = rN), lekin resources limited hote hain, isliye hum ek factor (1 − N/K) multiply karte hain. Jab N chhota hai, yeh factor ~1 hota hai (fast growth); jab N = K (carrying capacity) ho jaata hai, factor 0 ho jaata hai aur growth ruk jaati hai. Result ek beautiful S-shaped curve hai. Isko solve karke hum future predict kar sakte hain bina real experiment kiye — yahi modeling ki power hai.

Enzyme kinetics (Michaelis–Menten) mein hum dekhte hain ki reaction rate hamesha nahi badhti — enzyme saturate ho jaata hai, isliye v = Vmax[S]/(K_M+[S]). Aur predator-prey (Lotka–Volterra) mein do coupled equations milkar oscillations dikhate hain — jo emergent behaviour hai, ek single equation se nahi aati.

Yaad rakho recipe: variables likho, rates identify karo, balance equation banao (in minus out), rate law choose karo, solve karo, phir data se compare karo (Forecast-then-Verify). Sabse badi galti — sochna ki growth hamesha exponential rehti hai. Nahi! Real world mein resources limited hain, saturation aati hi hai.

Test yourself — Systems Biology & Frontiers