Systems and methods for using adverse event data to predict potential side effects

• US 9,235,686 B2
• Filed: 04/13/2012
• Issued: 01/12/2016
• Est. Priority Date: 01/06/2012
• Status: Active Grant

• Current Assignee: Molecular Health GmbH (Molecular Health AG)
• Original Assignee: Molecular Health GmbH (Molecular Health AG)
• Inventors: Jackson, David, Soldatos, Theodoros, Taglang, Guillaume, Zien, Alexander, Brock, Stephan
• Primary Examiner(s): Lin, Jerry
• Application Number:
  US13/446,912
• Publication Number:
  US 20130179138A1
• Time in Patent Office: 1,369 Days
• Field of Search: 702/19
• US Class Current: 1/1
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• CPC Class Codes: G16B 20/20 Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection
• CPC Class Codes: G16H 20/10 relating to drugs or medications, e.g. for ensuring correct administration to patients
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16H 70/40 relating to drugs, e.g. their side effects or intended usage

Abstract

The present disclosure describes systems and methods for predicting a likely side effect profile for even new, untested medications. A predicted side effect profile may be generated based on intersections of side effect profiles of other medications that affect the same or related molecular entities, such as the nearby target proteins, involve the same pathways, or are otherwise similarly related. To generate a predicted side effect profile for a new drug targeting a novel or previously un-targeted protein target, an analyzer may query an adverse event database for records pertaining to patients who have taken drugs or combinations of drugs that target or affect molecular entities in the vicinity of the novel target within a global molecular entity graph, and, in some embodiments, may retrieve a plurality of adverse event records and generate an intersection of side effects associated with related targets to predict likely side effects for the novel target.

Claims

1. A method for generating a predicted first side effect profile for a medication targeting a first target, comprising:

receiving, by an analyzer executed by a processor of a computing device from a user, an identification of a first medication and a first target, from a medication information database comprising medications and their associated targets, such that the first target is a target of the first medication;

identifying, by the analyzer from a global molecular entity graph, a second target functionally connected to the first target by;

generating a score for each of a plurality of targets responsive to interconnections with the first target in the global molecular entity graph, and

selecting the second target responsive to the second target having a highest score of the plurality of targets;

retrieving, by the analyzer from the medication information database, an identification of a second medication targeting the second target;

retrieving, by the analyzer from an adverse event database, a second side effect profile associated with the second medication;

generating, by the analyzer, the predicted first side effect profile for the first target based on the retrieved second side effect profile associated with the second medication; and

presenting, by a display module executed by the computing device, to the user, the predicted first side effect profile for the first target.

2. The method of claim 1, further comprising:

identifying a third target functionally connected to the first target;

retrieving, from the medication information database, an identification of a third medication targeting the third target;

retrieving, from the adverse event database, a third side effect profile associated with the third medication; and

wherein generating the predicted side effect profile for the first target comprises generating the predicted first side effect profile, based on an intersection of the second side effect profile and the third side effect profile.

3. The method of claim 1, wherein identifying a second target protein comprises selecting, via a shortest path algorithm, the second target protein from a plurality of target proteins functionally connected to the novel drug target in the graph.

4. The method of claim 1, wherein identifying a second target protein comprises selecting the second target protein from a plurality of target proteins functionally connected to the novel drug target in the graph, responsive to the second target protein having a highest number of nodal interconnections with the novel drug target of the plurality of target proteins.

5. The method of claim 1, wherein identifying a second target protein comprises selecting the second target protein from a plurality of target proteins functionally connected to the novel drug target in the graph, responsive to the second target protein having a fewest number of nodal interconnections to nodes not shared with the novel target of the plurality of target proteins.

6. The method of claim 1, wherein identifying a second target protein comprises selecting the second target protein responsive to a relationship between an organ associated with the second target protein and the novel drug target.

7. The method of claim 1, wherein identifying a second target protein comprises selecting the second target protein, responsive to the second target protein and novel drug target being included in a common pathway.

8. The method of claim 1, further comprising identifying a third target functionally connected to the first target having a second highest score of the plurality of targets;

retrieving, from the medication information database, an identification of a third medication targeting the third target;

retrieving, from the adverse event database, a third side effect profile associated with the third medication; and

wherein generating the predicted first side effect profile for the first target comprises generating the predicted first side effect profile, based on an intersection of the first side effect profile and the third side effect profile.

9. A system for generating a predicted first side effect profile for a medication targeting a first target, comprising:

a computing device comprising a processor executing

an analyzer, configured for;

receiving, from a user, an identification of a first medication and a first target, from a medication information database comprising medications and their associated targets, such that the first target is a target of the first medication;

identifying, from a global molecular entity graph, a second target functionally connected to the first target by;

generating a score for each of a plurality of targets responsive to interconnections with the first target in the global molecular entity graph, and

selecting the second target responsive to the second target having a highest score of the plurality of targets;

retrieving, from the medication information database, an identification of a second medication targeting the second target;

retrieving, from an adverse event database, a second side effect profile associated with the second medication; and

generating the predicted first side effect profile for the first target based on the retrieved second side effect profile associated with the second medication; and

a display module, configured for presenting, to the user, the predicted first side effect profile for the first target.

10. The system of claim 9, wherein the analyzer is further configured for:

identifying a third target functionally connected to the first target,

retrieving, from the medication information database, an identification of a third medication targeting the third target;

retrieving, from the adverse event database, a third side effect profile associated with the third medication; and

wherein generating the predicted first side effect profile for the first target comprises generating the predicted first side effect profile, based on an intersection of the second side effect profile and the third side effect profile.

11. The system of claim 9, wherein the analyzer is further configured for selecting, via a shortest path algorithm, the second target protein from a plurality of target proteins functionally connected to the novel drug target in the graph.

12. The system of claim 9, wherein the analyzer is further configured for selecting the second target protein from a plurality of target proteins functionally connected to the novel drug target in the graph, responsive to the second target protein having a highest number of nodal interconnections with the novel drug target of the plurality of target proteins.

13. The system of claim 9, wherein the analyzer is further configured for selecting the second target protein from a plurality of target proteins functionally connected to the novel drug target in the graph, responsive to the second target protein having a fewest number of nodal interconnections to nodes not shared with the novel target of the plurality of target proteins.

14. The system of claim 9, wherein the analyzer is further configured for selecting the second target protein responsive to a relationship between an organ associated with the second target protein and the novel drug target.

15. The system of claim 9, wherein the analyzer is further configured for selecting the second target protein, responsive to the second target protein and novel drug target being included in a common pathway.

16. The system of claim 9, wherein the analyzer is further configured for:

identifying a third target functionally connected to the first target having a second highest score of the plurality of targets;

retrieving, from the medication information database, an identification of a third medication targeting the third target;

retrieving, from the adverse event database, a third side effect profile associated with the third medication; and

wherein generating the predicted first side effect profile for the first target comprises generating the predicted first side effect profile, based on an intersection of the second side effect profile and the third side effect profile.