Which calculations does a cell culture bench do every day?

Our example lab's mycoplasma plate needs 240 µL of mix, and the scientist makes 264. The extra 24 µL is the overage, and it's the difference between filling the sixteenth well and running out at the fifteenth.

Most bench arithmetic is like that: simple, done in a hurry, and wrong in ways that only show later. Here are four calculations from one week in the lab our home page follows, with the numbers at every step: a mastermix, a cell count, a seeding plan and a doubling time. Then where each one goes wrong, and what the calculators in CCA Labs do and don't do.

How do you scale a PCR mastermix?

Multiply each component's volume per reaction by the number of reactions plus an overage, usually 10%, and top up with water. The overage covers what's lost to pipetting.

The plate is this month's mycoplasma screen: six samples, a positive control and a no-template control, each in duplicate, so 16 wells. Each well gets 15 µL of mix and 5 µL of sample. The mix doesn't include the sample, so the reaction volume for the calculation is 15 µL, not 20. With 10% on top, 16 × 1.1 = 17.6 reactions:

Component	µL per reaction	Scaled
2× qPCR master mix	10	176.0
Primer and probe mix, with internal control	2	35.2
Nuclease-free water	3	52.8
Total	15	264.0

Sixteen wells of 15 µL is 240 µL, which leaves 24 µL in the tube. In mechanical terms it's a tolerance stack. Each pipetting step loses a little, and the losses add up across sixteen wells. The overage is the margin you build in so the sum still fits.

How do you calculate cell concentration and viability?

Divide the live cells counted by the number of squares, multiply by the dilution factor and then by 10⁴, and you have live cells per mL. Viability is live cells divided by all cells counted.

The same lab's new HEK293 cell bank, counted on 17 September: the cells are mixed 1:1 with trypan blue, so the dilution factor is 2, and four large squares of a haemocytometer hold 400 live cells and 14 dead ones.

• Live cells per mL: 400 / 4 × 2 × 10⁴ = 2.00 × 10⁶
• Viability: 400 / 414 = 96.6%

The 10⁴ comes from the chamber. Each large square is 1 mm by 1 mm, 0.1 mm deep, so it holds 0.1 µL, or 10⁻⁴ mL. The classic slip is forgetting the trypan blue, which halves every count that day and makes the cells look slow for no reason.

How many cells should you seed per flask?

Multiply the target density, in cells per cm², by the growth area of the vessel. Then work out how much cell suspension holds that many cells and top up with medium.

On 14 September the bank's cells went into eight T175 flasks at 3.0 × 10⁴ cells per cm². A T175 has 175 cm² of growth area and a working volume of about 35 mL.

• Per flask: 3.0 × 10⁴ × 175 = 5.25 × 10⁶ cells
• Eight flasks with 10% overage: 5.25 × 10⁶ × 8 × 1.1 = 4.62 × 10⁷ cells
• From a suspension at 3.0 × 10⁶ per mL: 4.62 × 10⁷ / 3.0 × 10⁶ = 15.4 mL
• Total volume: 35 × 8 × 1.1 = 308 mL, so 292.6 mL of medium

Seeding by density rather than "a 1 in 5 split" makes flasks comparable from week to week, even when the culture you're splitting from was more or less confluent than usual.

How do you work out doubling time?

Doubling time is the hours between two counts, times ln 2, divided by the natural log of how many times the cells multiplied.

Those eight flasks held 4.20 × 10⁷ cells when they were seeded, not counting the overage left in the tube. Three days later, 72 hours, the harvest was 120 mL at 2.00 × 10⁶ per mL, which is 2.40 × 10⁸ cells.

doubling time = 72 × ln 2 / ln(2.40 × 10⁸ / 4.20 × 10⁷)
              = 49.91 / 1.743
              = 28.6 hours

Treat that number with some care. The formula assumes the cells grew at one steady rate the whole time. They didn't: they had a lag after seeding and slowed as the flasks filled up. So 28.6 hours is an average over the three days, and the real exponential rate was probably a bit faster. It's still worth recording every time, because the same calculation done the same way each passage shows you drift, and drift in growth rate is often the first sign that something about a culture has changed.

Where it goes wrong

None of these is hard. They go wrong for dull reasons.

• Units. Microlitres typed as millilitres, or cells per mL used as cells per cm². A calculator that shows the unit next to every field catches most of these.
• The forgotten factor. The trypan blue dilution, or the overage, or the fact that the sample isn't in the mix.
• Retyping. The number is right in the calculator and wrong in the notebook, because someone copied 35.2 as 32.5.
• Comparing unlike numbers. Doubling times from flasks seeded at different densities, or counted at different confluence, aren't the same measurement.

Bench Tools in CCA Labs

CCA Labs has 21 calculators in Bench Tools, which opens over whatever page you're on:

• General: a scientific calculator, unit converter, timers and stopwatches, and a code scanner that opens a sample or inventory item from its label.
• Chemistry & Prep: dilution (C1V1 = C2V2), serial dilution, mass and molarity, and the solution and mastermix builder used above.
• Molecular Biology: DNA and RNA quantification from A260, a multi-channel cell counter for concentration and viability, g-force to rotor speed, and protein parameters from a sequence.
• Assay Analysis: plate assay standard curves, assay statistics, the qPCR ΔΔCt method and cell culture (seeding by vessel, doubling time, projections and split ratios).
• Business & Money: margin, VAT, burn rate and runway, cost per sample, and rNPV.

One limit worth knowing before you rely on them. When you save a calculation, it goes into your own history in that tool, not into the experiment you're running. So the mastermix volumes still get typed into the step's record, and that's exactly the retyping problem above. We'd like to close it. The full list is on the all-in-one lab page, and you can create your lab to use them next to your own experiments.

The plate goes into the instrument at 11:15 and comes out at 12:45. Four cultures and the bank come back negative, the controls behave, and the fifth culture gives no signal at all, not even from the internal control. All of it goes into a record that lands on QA's desk that afternoon, which is the last part of the story: what a batch record should contain before you sign it.