Modelling a Biological System: Rate Equations, Parameters and Fit · seed 1 · A4, ink-friendly. The answer key prints on its own page for grown-ups.

Rate models you can defend from data

Science · Genetics & Evolution · ages 23-24
Name ______________________   Date ____________

  1. Why must you state the units of every parameter?

    • Journals require a units table
    • Terms only balance when units match across the line
    • Units make the code run faster

  2. What belongs on one line of a pathway rate model?

    • Inflows minus outflows for one molecule
    • The full pathway diagram redrawn
    • A list of every parameter value

  3. A model tracing old data perfectly is proven good, with no need to test on fresh data.

    Circle one:   True   False

  4. Your conclusion flips when one parameter shifts slightly. What does the sensitivity check say?

    • The conclusion leans entirely on that parameter
    • The model is fully identified
    • The units must be wrong

  5. With only sparse early growth data, which parameter stays poorly pinned?

    • Both are pinned equally
    • Growth speed
    • Capacity

  6. How do you judge fitted parameters honestly?

    • By the fit on the training points
    • By blind prediction on data held back
    • By counting the parameters

  7. Two parameters trade off so only their ratio is pinned. Which claim may you still defend?

    • A conclusion resting only on their ratio
    • Any prediction needing each value alone
    • The exact value of each parameter

  8. The full model disagrees with its solvable core in a simple limit. Where do you look first?

    • The blind test data
    • The crowding term only
    • The code or the units
LightMySky · lightmysky.comW1-mt_Al6sNjUOW5-s1

Answer key

For grown-ups. Fold this page away before handing over the rest.

Rate models you can defend from data W1-mt_Al6sNjUOW5-s1

  1. Terms only balance when units match across the line · Mismatched units mean the balance line cannot be right.
  2. Inflows minus outflows for one molecule · Each line balances a single pool: what enters minus what leaves.
  3. False · Perfect traces of old data often mean overfitting; blind prediction decides.
  4. The conclusion leans entirely on that parameter · High sensitivity means the conclusion stands or falls with that value.
  5. Capacity · Early data shows free growth; the ceiling only shows late.
  6. By blind prediction on data held back · Held-back data scores what the parameters learned, not memorized.
  7. A conclusion resting only on their ratio · Identifiability limits you to what the data actually pins down.
  8. The code or the units · Simple limits expose coding and unit slips before anything deeper.
Worksheet · LightMySky