MOI Calculator

Calculate MOI, expected infection rate, and virus/bacteria needed for infection experiments.

Number of infectious particles (virus/bacteria) (particles)

Number of target cells (cells)

Target MOI

Desired particles per cell

Total cells in experiment (cells)

For calculating particles needed

Please check your inputs and try again.

Why Poisson Rather Than Simple Division

MOI is the ratio of virus to cells, but infection is stochastic — particles land independently, so some cells receive several while others receive none. The distribution of hits per cell is Poisson.

P(k hits) = e^−MOI × MOI^k / k! % infected = (1 − e^−MOI) × 100

MOI	% infected	% uninfected	% of infected with >1
0.1	9.5%	90.5%	5%
0.3	25.9%	74.1%	14%
1.0	63.2%	36.8%	42%
3.0	95.0%	5.0%	84%
10	99.995%	0.005%	~100%

Two results surprise people. MOI 1 infects only 63%, not all cells — more than a third receive nothing. And reaching 99.995% requires MOI 10, by which point essentially every infected cell has taken multiple particles.

Why Screens Use Low MOI

In a pooled CRISPR or shRNA screen each cell must carry one perturbation, so its phenotype can be attributed to that guide. A cell with two integrations gives an uninterpretable result and adds noise.

This is why screens target 20–30% infection despite discarding most cells. At MOI 0.3 about 14% of infected cells carry more than one construct; at MOI 1 it would be 42%, which corrupts the data.

The cost is library coverage: fewer infected cells means fewer per guide, so starting cell numbers must rise to maintain 500–1,000 cells per guide after selection.

Functional Titre Is Cell-Type Specific

Factor	Effect on effective MOI
Target cell type	Can differ by orders of magnitude
Receptor expression	Determines entry efficiency
Polybrene or similar	Improves adsorption
Spinoculation	Raises effective MOI substantially

Titre must be measured functionally on the actual target cells. A preparation giving 10⁸ TU/mL on HEK293T may achieve a small fraction of that on primary T cells, so applying the HEK-derived number produces a much lower MOI than intended.

Worked Examples

Example 1: 1M virus + 100K cells

MOI=10, P(0 hits)=e^-10=0.005%

Result: 99.995% cells infected

High MOI for synchronous infection

Example 2: MOI=1 for CRISPR screen

P(infected)=63.2%, P(0 hit)=36.8%

Result: MOI=0.3 for lentivirus: 26% infected (1 integration/cell)

Low MOI ensures single integration per cell

Example 3: Why MOI 0.3 for screens

MOI 0.3 versus MOI 1

Result: 14% versus 42% multiple integrations

At the higher MOI nearly half of infected cells carry more than one guide, and their phenotypes cannot be attributed to either construct.

Example 4: Titre on the wrong cells

10⁸ TU/mL measured on HEK293T, applied to primary T cells

Result: Effective MOI far lower

Primary cells transduce much less efficiently. The intended MOI is not achieved, and infection falls well short of target.

Common Mistakes

⚠️

Assuming MOI 1 infects every cell

Poisson statistics give 63.2%. More than a third of cells receive no particle at all at that ratio.

⚠️

Using high MOI for pooled screens

Multiple integrations per cell make phenotypes uninterpretable. Screens target 20–30% infection so most infected cells carry one construct.

⚠️

Applying a titre from a different cell type

Transduction efficiency varies by orders of magnitude between cell types. Functional titre must be measured on the actual target.

⚠️

Forgetting coverage when lowering MOI

Fewer infected cells means fewer per guide. Starting numbers must increase to keep 500–1,000 cells per guide after selection.

Frequently Asked Questions

Why Poisson distribution?

Viral infection is random — each particle independently infects any cell. If average is λ particles/cell, P(k infections) = e^(-λ)×λ^k/k!. For λ=MOI: P(0)=e^(-MOI), P(≥1)=1-e^(-MOI).

MOI in CRISPR screens?

Use MOI=0.3 for lentiviral CRISPR screens: P(2+ integrations)=4%, P(1 integration)=26%, P(0)=74%. Low MOI ensures most infected cells get exactly one sgRNA — critical for screen interpretation. Infect large number, select only infected cells.

Why doesn't MOI 1 infect all cells?

Because particles distribute randomly. Poisson statistics give e⁻¹ = 36.8% of cells receiving nothing, so only 63.2% are infected.

What MOI should a pooled screen use?

Low enough for 20–30% infection, typically MOI 0.3. This keeps most infected cells at a single integration so phenotypes are interpretable.

How do I calculate MOI for a target infection rate?

Take the negative natural log of the uninfected fraction. For 30% infection, MOI = −ln(0.70) = 0.357.

Why measure titre on the target cell type?

Because transduction efficiency varies enormously between cell types. A titre from HEK293T can overstate performance on primary cells by orders of magnitude.