Systems and methods for antibody engineering

A method for determining a measure of confidence for a descriptor in a first sequence activity relationship for an antibody of interest, the method comprising:

• (a) identifying a plurality of positions in said antibody of interest and, for each respective position in said plurality of positions, one or more substitutions for the respective position, wherein the plurality of positions and the one or more substitutions for each respective position in the plurality of positions collectively define an antibody sequence space;
• (b) selecting a first plurality of variants of the antibody of interest, thereby forming a variant set, wherein said variant set comprises a subset of said antibody sequence space, wherein each respective variant in the first plurality of variants has one or more substitutions in the amino acid sequence of the antibody of interest;
• (c) modeling, using a suitably programmed computer, a plurality of sequence-activity relationships between (i) a respective plurality of descriptors, wherein each descriptor in the respective plurality of descriptors is a descriptor for one or more corresponding substitutions at one or more corresponding positions of the antibody of interest in the variant set and (ii) a plurality of quantitative measures of a property, wherein each quantitative measure in the plurality of quantitative measures is a measurement of the property exhibited by a respective variant in the variant set, the plurality of sequence-activity relationships including the first sequence activity relationship, and wherein
• each respective sequence-activity relationship in said plurality of sequence-activity relationships includes at least one descriptor that is found in another sequence-activity relationship in said plurality of sequence-activity relationships;
• each respective sequence-activity relationship in said plurality of sequence-activity relationships is either
• (A) between (i) a subset of the plurality of descriptors and (ii) the plurality of quantitative measures or
• (B) between (i) the plurality of descriptors and (ii) a subset of the plurality of quantitative measures;
• each descriptor in the plurality of descriptors is weighted by a corresponding weight in a plurality of weights, and wherein said modeling comprises determining, for each respective weight in the plurality of weights, a respective first value and a respective second value for the respective weight using sequence-activity relationships in the plurality of sequence-activity relationships that include the descriptor corresponding to the weight, wherein the respective second value is a measure of confidence in the respective first value.