Side effects prediction using co-associating bioattributes

摘要

A bioinformatics method, software, database and system for predicting side effects resulting from predisposition modification are presented in which a set of attributes is used to modify an individual's attribute profile to generate a modified attribute profile. The modified attribute profile is compared against a predetermined set of attribute combinations that are statistically associated with one or more candidate side effect attributes. Attribute combinations from the set that also occur in the modified attribute profile are identified, and those candidate side effect attributes that are statistically associated with the identified attribute combinations are stored as a set of predicted side effect attributes resulting from modification of the individual's attribute profile.

主权项

1. A bioinformatics computer for predicting future side effects that result from modifying an attribute profile of an individual, said computer including a data input device, a processor, a memory, and data output device, wherein a method performed on said computer is characterized as comprising:
a) said processor accessing a previously modified attribute profile contained in said memory, wherein the previously modified attribute profile comprises an attribute profile of said individual that has been previously modified as to a set of pre-existing attributes associated with that individual, wherein the identity of the individual is masked or anonymized; b) said processor accessing a dataset in said memory containing attribute combinations and corresponding statistical results that indicate the strength of association of each of the attribute combinations with one or more candidate side effect attributes, wherein each of the attribute combinations comprises at least one pangenetic attribute and at least one non-pangenetic attribute, and wherein the one or more candidate side effect attributes comprise pangenetic attributes and non- pangenetic attributes that do not occur in the previously modified attribute profile, the candidate side effect attributes being filtered to exclude user specified attributes; c) said processor identifying attribute combinations contained in the dataset that also occur in the previously modified attribute profile; d) transmitting through said output device one or more candidate side effect attributes that are statistically associated with the identified attribute combinations as output, to generate a set of predicted cumulative future side effect attributes that result from modifying the attribute profile of the individual, wherein the candidate side effect attributes transmitted as output are transmitted as a rank-ordered tabulation containing corresponding statistical results that indicate:
(i) the strength of association of the candidate side effect attributes in the tabulation with the identified attribute combinations, and(ii) the individual's potential for association with each of the predicted cumulative future side effect attributes; e) updating the modified attribute profile with the predicted cumulative future side effect attributes; and f) repeating steps (a)-(e) until no further attribute combinations can be identified in step (c); g) compiling the predicted cumulative future side effect attributes into at least two subgroups corresponding to desirable and undesirable side effects.