What it does
Cells in steady growth double at a constant pace, so the number of cells rises as N = N₀ × 2^(t / Td), where Td is the doubling time. This page works out Td from your own counts and then uses it to plan the next split.
- Two counts. Td = (t₂ − t₁) × ln 2 / ln(N₂ / N₁). Four times as many cells after 48 hours is two doublings, so Td = 24 h.
- Three or more counts. On a log scale steady growth is a straight line. The page fits ln(count) against time by least squares: the slope is the growth rate r (per hour), Td = ln 2 / r, and the slope’s standard error gives a 95 % interval for Td. R² shows how straight the line is.
- Generations are the doublings over the fitted span: r × time ÷ ln 2.
- Between each pair of counts the table shows a doubling time for every interval, so you can see a slow start (lag) or growth levelling off (plateau) and untick those counts.
Plan the next passage
The planner runs the same law backwards. To have a target confluence on a given day, seed the target number of cells divided by 2^((hours − lag) / Td). It tells you:
- Cells to seed per vessel, as cells, cells per cm² and percent confluence.
- Split ratio, by area: the density now divided by the seeding density (1 : 8 means the cells from one flask spread over the area of eight).
- Volumes: how much cell suspension and how much medium to put in each vessel, how many cells your vessels need and how many you could seed.
- The next passages: dates, the split each time, population doublings since today and how many cells you would have if you kept every cell, with a plot of the predicted confluence.
How to use it
- Counts. Type the time and count for each sample, or paste two columns from a spreadsheet. Counts can be cells/mL, cells per well or confluence, as long as they are all measured the same way. Choose hours or days.
- Fit. Untick counts from the lag or the plateau. Read the doubling time, its interval and R².
- Seed. The doubling time is copied into the planner; you can type your own. Pick the vessels the cells are in now and either their confluence or your count (cells/mL and the volume they are resuspended in). Then pick the new vessels, how many, the target confluence and when you want them ready.
- Schedule. Read the plan and the next passages, and download the counts or the schedule as CSV, or the whole report as an image, PDF or Word file.
Worked example
The example uses made-up counts in cells/mL: 100,000 at 0 h, 180,000 at 24 h, 410,000 at 48 h, 820,000 at 72 h, 1,600,000 at 96 h and 1,900,000 at 120 h. The last count is left unticked because growth has levelled off (its own interval shows a doubling time of 96.8 h). The fit through the other five gives a doubling time of 23.56 h (0.9816 days), a growth rate of 0.02942 per hour (0.7062 per day), a 95 % interval of 21.77 to 25.67 h, R² = 0.9980 and 4.075 generations in 96 hours (16.85-fold). Tick the last count too and the doubling time drifts up to 26.5 h with a warning that growth slows down.
For the passage, a T75 flask (75 cm²) is 80 % confluent. At 1 × 10⁵ cells per cm² for a full layer that is about 6,000,000 cells, resuspended in 5 mL (1,200,000 cells/mL). To have two T75 flasks 80 % confluent in 3 days, with no lag, the cells need 3.056 doublings, so seed 721,400 cells per T75 flask (9,619 cells/cm², about 9.619 % confluent). The split ratio is 1 : 8.317. Put 601 µL of suspension and 14.4 mL of medium in each flask. The two flasks take 1,443,000 cells, and you have enough for 8.
Tips
- Count at the same time of day, in the same way, from the same vessels. Mix well before every count.
- Seed a few vessels and count one each day: a series of four to six counts gives a much steadier doubling time than two.
- Freshly split cells often take half a day to a day to start dividing. Put that in Lag after seeding.
- Growth slows as cells fill the surface, so plans that aim above about 90 % arrive late. Most lines are best split before they are fully confluent.
- Very sparse cells may grow slowly or not at all; the page warns below 5 % confluence.
FAQ
What counts as one doubling? One population doubling: the number of cells goes from N to 2N. Generations, population doublings and doublings are the same thing here.
Why does my two-count doubling time differ from the fit? Two counts fit exactly and carry every error of both counts. A fit through several counts averages those errors out and gives an interval.
Can I use confluence instead of counts? Yes, while the cells are well below confluence. Near a full layer, confluence stops rising exponentially.
Are the vessel areas exact? No. They are typical rounded values; check your own flasks and type the real area and medium volume.
Is anything sent or saved? No. Everything is worked out in your browser and nothing is stored.
Where the numbers come from
- Exponential growth, N = N₀ × 2^(t / Td), the two-count formula Td = (t₂ − t₁) × ln 2 / ln(N₂ / N₁) and the least-squares fit of ln(count) against time are general science knowledge, written by fibub.
- Checked in the tests against an independent NumPy least-squares script with its own Student-t quantile, on the example and four random growth series (doubling time, its 95 % interval, R², generations), and the passage planner against hand working and the same script.
- Vessel areas and medium volumes are typical rounded values chosen by fibub, not any maker's table; 1 × 10⁵ cells per cm² at 100 % confluence is a round starting value. All are editable.
- The example counts are made up for this tool.
Standard scientific values, such as constants, atomic weights and melting points, are facts found in every textbook; we name where ours were checked. If you are an author or publisher and want a value, a credit or a table changed or removed, tell us and we will act quickly.
Last updated 2026-11-09.