Question 1

How do I calculate specific growth rate from OD data?

Accepted Answer

To calculate the specific growth rate (µ) from OD data, first identify the exponential (log) phase of your growth curve where OD increases exponentially. Plot ln(OD) versus time — the exponential phase will appear as a straight line. The slope of this line equals the specific growth rate µ (in h⁻¹). You can then calculate the doubling time as td = ln(2)/µ. For more accurate results, use at least 4–6 data points within the exponential phase and ensure your OD values are within the linear range of your spectrophotometer (typically OD < 0.4, or dilute samples appropriately).

Question 2

What is the difference between logistic and Gompertz models?

Accepted Answer

Both the logistic and Gompertz models describe sigmoidal (S-shaped) growth curves, but they differ in symmetry and biological interpretation. The logistic model is symmetric around its inflection point (at K/2), while the Gompertz model is asymmetric with the inflection point at K/e (~37% of K). The Gompertz model typically fits microbial growth data better because it explicitly accounts for lag time (λ) and the gradual deceleration phase. In microbiology, the modified Gompertz model is the most widely used equation for describing batch growth curves.

Question 3

Why does my growth curve not fit well?

Accepted Answer

Poor curve fitting can result from several issues: (1) Insufficient data points, especially during the transition phases. Aim for at least 8–12 points across the full growth curve. (2) Noisy or inconsistent measurements — ensure proper blanking and mixing before OD readings. (3) Diauxic growth (two growth phases) which cannot be described by a single sigmoidal model. (4) OD values outside the linear range of the spectrophotometer (>0.4 OD without dilution). (5) The wrong model for your data — try all three models and compare R² values. (6) Growth inhibition, substrate limitation, or pH shifts causing atypical growth patterns.

Question 4

How many data points do I need for accurate fitting?

Accepted Answer

For reliable growth curve fitting, you need a minimum of 6–8 data points, but 12–20 points are recommended for accurate parameter estimation. The points should be distributed across all growth phases: at least 2–3 points in the lag phase, 4–6 points in the exponential phase (this is critical for accurate µ estimation), and 2–3 points in the stationary phase. More frequent sampling during the exponential phase improves the accuracy of µmax estimation. For the Gompertz model, having clear lag and stationary phase data is essential for estimating λ and K parameters.

Question 5

Should I use OD600 or cell counts for growth curves?

Accepted Answer

Both OD600 and direct cell counts are valid but serve different purposes. OD600 is faster, non-destructive, and ideal for real-time monitoring, but it becomes non-linear above ~0.4 OD and measures turbidity (not actual cell number). It can be affected by cell morphology changes and doesn't distinguish live from dead cells. Direct cell counts (hemocytometer, flow cytometry, or automated counters) give true cell concentrations and can assess viability (with trypan blue or PI staining). For accurate µ calculations, cell counts are more reliable. For routine monitoring, OD600 is practical — just ensure you dilute samples into the linear range and establish an OD-to-cell-count calibration curve for your organism.

Question 6

How do I identify the exponential phase?

Accepted Answer

The exponential (log) phase is identified by plotting your data on a semi-log scale (ln of cell concentration vs. time). During exponential growth, this plot will show a straight line. The phase typically begins after the lag phase when cells have adapted to the medium, and ends when nutrients become limiting or inhibitory metabolites accumulate. In this tool, you can use the log-scale toggle to visually identify the linear region, or use the manual phase selector to set the start and end times. The tool also auto-detects the exponential phase by finding the region with the highest sustained slope in the ln-transformed data.

Growth Curve Fitter

About Growth Curve Fitting

Further reading