01 / 20

Interpreting Graphs &
Multi-Panel Figures

Learn to decode line plots, bar charts, and multi-panel figures commonly found in biochemistry research: from enzyme kinetics to gene expression data.

Biochemistry

Why Graphs Matter in Biochemistry

Biochemistry relies on quantitative data: enzyme reaction rates, protein concentrations, absorbance spectra, and gene expression levels. Graphs transform raw numbers into visual patterns that reveal trends, thresholds, and relationships far faster than tables alone.

Core Graph Vocabulary

Independent Variable
The variable the experimenter controls: plotted on the x-axis (e.g., substrate concentration \([S]\)).
Dependent Variable
The measured outcome: plotted on the y-axis (e.g., reaction velocity \(v\)).
Controls
Baseline conditions shown as a separate curve or bar group for comparison.

Common Graph Types in Biochemistry

Line Plot
Continuous data over time or concentration: enzyme kinetics, growth curves.
Bar Chart
Discrete comparisons between conditions: gene expression fold-change, protein levels.
Scatter Plot
Correlation between two variables: Lineweaver-Burk double-reciprocal plots.
Dose-Response
Sigmoidal curves relating ligand concentration to biological effect: \(\text{EC}_{50}\), \(\text{IC}_{50}\).

Anatomy of a Well-Labeled Axis

2 4 6 8 25 50 75
\([S]\) (mM)
\(v\) (µmol/min)
 Title: Effect of [S] on Reaction Rate Units are essential!

Reading a Michaelis-Menten Curve

\([S]\) (substrate concentration)
\(v\) (reaction velocity)
\(V_{\max}\)
\(V_{\max}/2\)
\(K_m\)
 initial slope

Linear vs. Logarithmic Scales

Linear Scale
Equal spacing between values: 0, 10, 20, 30. Best for data spanning a narrow range.
Log Scale
Each tick is a power of 10: \(10^1, 10^2, 10^3\). Essential when data spans several orders of magnitude (e.g., dose-response curves).

Reading Multi-Panel Figures

A Time (min) Absorbance WT Mutant B Condition Activity (%) WT Mut A Mut B

Step-by-Step: Decoding Multi-Panel Figures

1
Read the figure legend first. It explains what each panel shows, what the symbols mean, and the number of replicates.
2
Check axes and units. Each panel may have different scales. Don't assume Panel B uses the same y-axis as Panel A.
3
Identify controls. Find the baseline (WT, vehicle, untreated) in every panel before comparing treatments.
4
Look for error bars. Overlapping error bars suggest no significant difference; non-overlapping bars suggest a real effect.
5
Connect across panels. Ask: what story do panels A, B, C tell together that no single panel tells alone?

Error Bars and Statistical Markers

SD (Standard Deviation)
Shows the spread of raw data. Larger SD means more variability among measurements.
SEM (Standard Error)
Shows precision of the mean estimate. \(\text{SEM} = \frac{\text{SD}}{\sqrt{n}}\). Smaller than SD by definition.
95% CI
Range where the true mean likely falls. If two CIs don't overlap, the difference is typically significant.

Dose-Response Sigmoidal Curves

log \([Drug]\) (M)
 % Response 0 25 50 75 100 -9 -8 -7 -6 -5
\(\text{EC}_{50}\) A
\(\text{EC}_{50}\) B
 50% Drug A Drug B

Worked Example: Enzyme Inhibition

\([S]\) (mM)
\(v\) (µmol/min)
\(V_{\max}\)
 No inhibitor + Competitive inhibitor

Worked Example: Western Blot + Quantification

A Western Blot Ctrl Drug KO Target Actin Normalized to actin B Densitometry Condition Relative level 0 0.5 1.0 Ctrl Drug KO ★★

Connecting Panels: Building the Story

A typical biochemistry figure tells a multi-layered story. Consider a three-panel figure studying a new enzyme inhibitor:

A
In vitro assay: Dose-response curve showing the inhibitor reduces enzyme activity with an \(\text{IC}_{50}\) of \(2.5\,\mu\text{M}\).
B
Cell-based assay: Western blot showing the downstream product decreases with treatment, confirming the pathway is blocked in living cells.
C
Selectivity: Bar chart comparing inhibition across five related enzymes: only the target enzyme is significantly affected.

Common Mistakes in Graph Interpretation

✕ Ignoring Error Bars
Concluding a difference is "real" when error bars overlap heavily. Always check for statistical significance markers.
✕ Confusing Log and Linear
Interpreting equal spacing on a log axis as equal differences. A jump from \(10^{-8}\) to \(10^{-7}\) is a 10-fold change.
✕ Truncated Y-Axis
A y-axis starting at 90 instead of 0 can make a 5% change look dramatic. Always check where the axis starts.
✕ Correlation as Causation
Two variables trending together on a scatter plot does not prove one causes the other. Look for controlled experiments.

Graph Types: Quick Reference

Graph Type Best For Key Feature to Read Watch Out For
Line Plot Continuous trends over time or concentration Slope, plateau, inflection Interpolation between sparse points
Bar Chart Comparing discrete conditions Height differences + error bars Truncated y-axis
Scatter Plot Correlation between two variables Trend line, \(R^2\) value Outliers distorting the fit
Dose-Response Drug potency and efficacy \(\text{EC}_{50}\), max response Log-scale x-axis compression

Question 1 of 5

An enzyme kinetics graph shows two curves: one without inhibitor reaches a \(V_{\max}\) of 100 µmol/min, and one with inhibitor reaches a \(V_{\max}\) of 50 µmol/min. Both have the same \(K_m\). What type of inhibition is this?

Question 2 of 5

On a dose-response curve with a log-scale x-axis, Drug X has an \(\text{EC}_{50}\) of \(10^{-8}\) M and Drug Y has an \(\text{EC}_{50}\) of \(10^{-6}\) M. Both reach the same maximum response. Which statement is correct?

Question 3 of 5

A bar chart shows protein expression levels. The control bar height is 1.0 with an error bar (SEM) of \(\pm 0.1\). The treatment bar is 1.15 with an SEM of \(\pm 0.12\). No significance asterisk is shown. What can you conclude?

Question 4 of 5

A multi-panel figure includes: Panel A (a time-course showing product formation) and Panel B (a bar chart of final product at 60 min). In Panel A, Condition X plateaus at 30 min while Condition Y keeps rising. In Panel B, both conditions show similar bar heights. What is the best interpretation?

Question 5 of 5

A researcher presents a scatter plot of body weight vs. blood glucose in 200 patients with an \(R^2 = 0.85\). They conclude: "Increased body weight causes elevated blood glucose." What is the flaw in this reasoning?