Interactive-aware clustering of stable states

Patent metadata

Field	Value
Publication number	US-11515005-B2
Application number	US-201916284470-A
Country	US
Kind code	B2
Filing date	Feb 25, 2019
Priority date	Feb 25, 2019
Publication date	Nov 29, 2022
Grant date	Nov 29, 2022

Abstract

Analysis of genetic disease progression may be provided. Data about a set of molecular status may be received. A dynamic prediction model of molecular interactions may be provided over time. The molecular statuses of the set over time may be determined using the dynamic prediction model. The determined molecular statuses may be clustered by applying an interaction-aware metric for the analysis of the genetic disease progression.

First claim

What is claimed is:

1. A computer-implemented method for an analysis of genetic disease progression, said method comprising:
receiving data of a set of molecular statuses;
providing a dynamic prediction model of molecular interactions over time;
determining said molecular statuses of said set over time using said dynamic prediction model; and
clustering said determined molecular statuses by applying an interaction-aware metric for said analysis of said genetic disease progression, the interaction-aware metric including a framework of determining distances considering a network topology of interacting genes, wherein the interaction-aware metric that determines the distances considers multidimensional molecular interaction representing at least an influence of a concentration of a gene on another gene, wherein the clustering considers underlying molecular interactions in addition to distances between the molecular statuses.

2. The method according to claim 1 , wherein said dynamic prediction model is selected out of said group comprising a Boolean model, a set of ordinary differential equations, a set of stochastic differential equations, a set of partial differential equations.

3. The method according to claim 1 , wherein said set of molecular statuses is represented by measurement values of a patient, a cell line or drug perturbations.

4. The method according to claim 1 , wherein said dynamic prediction model and related parameters are downloadable from a remote database.

5. The method according to claim 1 , wherein said interaction-aware metric is at least in parts based on a Laplacian matrix, a deformed Laplacian matrix, a symmetric Laplacian matrix, a random walk Laplacian matrix and/or a Vicus matrix.

6. The method according to claim 1 , wherein said analysis of said genetic disease progression comprises
determining a personalized therapy.

7. The method according to claim 1 , also comprising
determining a disease sub-type based on said clustering.

8. The method according to claim 1 , wherein said clustering is based on at least one algorithm selected out of said group k-means, DBSCAN, and hierarchical clustering.

9. The method according to claim 1 , wherein as distance measure during said clustering a Manhattan distance or an Euclidean distance is determined.

10. The method according to claim 9 , wherein said distance determination is also a function of said interaction-aware metric.

11. A disease analysis system for an analysis of genetic disease progression, said system comprising
a receiving unit adapted for receiving data about a set of molecular statuses;
a prediction module adapted for a dynamic prediction model of molecular interactions over time;
a determination unit adapted for determining said molecular statuses of said set over time using said dynamic prediction model; and
a clustering module adapted for clustering said determined molecular statuses by applying an interaction-aware metric to analyze genetic disease progression, the interaction-aware metric including a framework of determining distances considering a network topology of interacting genes, wherein the interaction-aware metric that determines the distances considers multidimensional molecular interaction representing at least an influence of a concentration of a gene on another gene, wherein the clustering considers underlying molecular interactions in addition to distances between the molecular statuses.

12. The system according to claim 11 , wherein said dynamic prediction model is selected out of said group comprising a Boolean model, a set of ordinary differential equations, a set of stochastic differential equations, a set of partial differential equations.

13. The system according to claim 11 , wherein said data of said set of molecular statuses is represented by measurement values of a patient, a cell line or drug perturbations.

14. The system according to claim 11 , also comprising
a download module adapted for downloading said dynamic prediction model and related parameters from a remote database.

15. The system according to claim 11 , wherein said interaction-aware metric is at least in parts based on a Laplacian matrix, a deformed Laplacian matrix, a symmetric Laplacian matrix, a random walk Laplacian matrix and/or a Vicus matrix.

16. The system according to claim 11 , also comprising
personalization determination module adapted for a determination of a personalized therapy as part of said analysis of said genetic disease progression.

17. The system according to claim 11 , wherein said clustering module unit is also adapted for
determining a disease sub-type.

18. The system according to claim 11 , wherein said clustering module unit is adapted for performing at least one algorithm selected out of said group k-means, DBSCAN, and hierarchical clustering.

19. The system according to claim 11 , wherein as distance measure during said clustering a Manhattan distance or an Euclidean distance is determined.

20. A computer program product for an efficient analysis of genetic disease progression, said computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, said program instructions being executable by a computer to cause said computer to:
receive data about a set of molecular statuses;
provide a dynamic prediction model of molecular interactions over time;
determine said molecular statuses of said set over time using said dynamic prediction model; and
cluster said determined molecular statuses by applying an interaction-aware metric to analyze genetic disease progression, the interaction-aware metric including a framework of determining distances considering a network topology of interacting genes, wherein the interaction-aware metric that determines the distances considers multidimensional molecular interaction representing at least an influence of a concentration of a gene on another gene, wherein clustering considers underlying molecular interactions in addition to distances between the molecular statuses.

Assignees

• IBM

Inventors

• Purandare Mitra
• Manica Matteo
• Polig Raphael
• Rodriguez Martinez Maria

Classifications

• G16B20/50 Primary

  Mutagenesis · CPC title

• G16B40/30 Primary

  Unsupervised data analysis · CPC title

• G06F18/231

  Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram · CPC title

• G06F18/2323

  based on graph theory, e.g. minimum spanning trees [MST] or graph cuts · CPC title

• G06F18/21355

  nonlinear criteria, e.g. embedding a manifold in a Euclidean space · CPC title