A method of determining aggregation rate data predicting an aggregation rate of a polypeptide defined by an amino acid sequence, the method comprising determining a hydrophobicity value, a charge value, and at least one shape propensity value for said sequence; identifying one or more aggregation-influencing patterns within said sequence; determining a pattern value for the sequence responsive to said identifying; and determining said aggregation rate data by determining a weighted combination of said hydrophobicity value, said charge value, said at least one shape propensity value, said pattern value and at least one factor extrinsic to said amino acid sequence.