Steps involved:

- The data is subtracted by the mean of the wild type.
- The sum of these zero-centered replicates plus the mean of the wild type is the theoretical mean assuming that these values are additive (i.e. have no epistasis).
- The theoretical standard error is calculated by taking the square root of the sum the variances of the sets of replicates divided by number of replicates.
- If the epistatic effect is beneficial, it is positive (+), if detrimental negative (-). Depending on the contributing values, there are different types of epistasis.
- When the beneficial or detrimental gains of the base mutations simply stack by summation, (ie. without epistasis), this is
**additive**. - When base mutations with the beneficial and detrimental gains are combined, but the result is greater than simply additive, this us
**sign epistasis**. - When the beneficial or detrimental gains of the base mutations stack more than expected, this is
**magnitude epistasis**. - When the beneficial or detrimental gains of the base mutations result in the reverse effect, this is
**reciprocal sign epistasis**.