USER INTERFACE, SYSTEM, AND METHOD FOR COHORT ANALYSIS

• US 20200135303A1
• Filed: 10/31/2019
• Published: 04/30/2020
• Est. Priority Date: 10/31/2018
• Status: Active Grant

• Current Assignee: Tempus Labs, Inc.
• Original Assignee: Tempus Labs, Co.
• Inventors: Barber, Mathew
• Granted Patent: US 11,521,710 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G16B 25/10 Gene or protein expression profiling; Expression-ratio estimation or normalisation
• CPC Class Codes: G16B 40/20 Supervised data analysis
• CPC Class Codes: G16B 40/30 Unsupervised data analysis
• CPC Class Codes: G16B 45/00 ICT specially adapted for bioinformatics-related data visualisation, e.g. displaying of maps or networks
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 20/00 ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A system and method that receive a distance matrix for multiple patients and a patient of interest, assign a radial distance value between the patient of interest and the other patients based on the distance matrix value for each of the multiple patients, generate an angular distance value between the multiple patients based at least in part on a measure of similarity between each patient, and minimize a cost function based at least in part on the angular distance value between each patient and each other patient. Minimizing the cost function may include calculating a patient contribution to the cost function for a plurality of angular distance values and selecting the angular distance value with the smallest patient contribution. The processor also may be configured to generate and display a radar plot based on the assigned radial distance value and generated angular distance value of each patient.

Claims

1. ) A method, comprising:

receiving a first dataset, the first dataset having at least a first, second, and third member, where the first, second, and third members are matched to a plurality of identifiers;

receiving a second dataset, the second dataset having a plurality of importance signifiers, where one or more of the plurality of importance signifiers are matched to one or more of the plurality of identifiers of the first dataset;

generating a third dataset based at least in part from the first and second dataset, the third dataset having the plurality of identifiers of the first dataset matched with a plurality of scaled importance factors;

generating, by a processing device, a fourth dataset based at least in part on the first, second, and third datasets, the fourth dataset having a measure of similarity between the first member and the second member, a measure of similarity between the first member and the third member, and a measure of similarity between the second member and the third member;

receiving a selection of the first member from a user;

identifying a first angular assignment for the second member and a second angular assignment for the third member, wherein the first and second angular assignments minimize a sum of squared error function and the second angular assignment indicates a similarity between the second member and the third member; and

displaying, via a display operatively connected to the processing device, the first member at a focal point of a graphical user interface on a display, the second member at a first radial distance based on the measure of similarity between the first and second member and the first angular assignment, and the third member at a second radial distance based on the measure of similarity between the first and third member and the second angular assignment.

2. ) The method of claim 1, wherein the first member is an oncology patient, wherein a first identifier of the plurality of identifiers is a first genetic mutation, and wherein the plurality of importance signifiers include evidence of FDA-approved therapy for each gene and evidence of correlation to causing cancer for each gene.

3. ) The method of claim 2, further comprising:

processing first and second datasets and extracting clusters of interest.

4. ) The method of claim 2, further comprising calculating the measure of similarity between the first member and the second member by:

assigning a first value based at least in part on a number of gene mutations the first and second member have in common; and

modifying the first value by a second value based at least in part on a number of gene mutations the first and second member do not have in common.

5. ) The method of claim 1, wherein a first identifier of the plurality of identifiers is a normalized gene expression count exceeding a threshold value for a first gene.

6. ) The method of claim 5, wherein the plurality of importance signifiers includes the first gene being part of one or more first genetic pathways.

7. ) The method of claim 6, wherein the plurality of identifiers includes normalized gene expression counts exceeding a threshold value for at least a second gene.

8. ) The method of claim 7, wherein the plurality of importance signifiers includes the at least a second gene being part of at least one of the one or more first genetic pathways.

9. ) The method of claim 5, wherein the graphical user interface includes a selectable region configured to permit the user to select a first gene from among a plurality of genes or a genetic pathway from among a plurality of genetic pathways.

10. ) A system, comprising:

a display; and

a processor, the processor operatively connected to the display, the processor to;

receive a distance matrix for a plurality of patients and a patient of interest;

assign a radial distance value between the patient of interest and the other patients based on the distance matrix value for each patient of the plurality of patients;

generate an angular distance value for each patient and each other patient of the plurality of patients based at least in part on a measure of similarity between each patient;

minimize a cost function based at least in part on the angular distance value between each patient and each other patient of the plurality of patients, wherein minimizing the cost function comprises;

calculating a patient contribution to the cost function for a plurality of angular distance values; and

selecting the angular distance value with the smallest patient contribution;

generate a radar plot based at least in part on the assigned radial distance value and generated angular distance value of each patient; and

cause the display to display the radar plot.

11. ) The system of claim 10, wherein the distance matrix is derived from a comparison of one or more genetic mutations in the patient of interest and the plurality of patients.

12. ) The system of claim 11, wherein the measure of similarity is derived from a comparison of the one or more genetic mutations among the plurality of patients.

13. ) The system of claim 10, wherein the patient of interest is an oncology patient, wherein the distance matrix is derived from a comparative analysis of cancer type as between the patient of interest and the plurality of patients.

14. ) The system of claim 13, wherein the measure of similarity is derived, at least in part, from responses to treatment for each of the plurality of patients.

15. ) The system of claim 14, wherein responses include an indication of a type of treatment, favorable results to a given therapy, a percentage of remission after one or more time periods, and/or an indication of an existence or lack of adverse side effects.

16. ) A method, comprising:

receiving a first dataset, the first dataset having at least a first, second, and third member, where the first, second, and third members are matched to respective gene expression counts;

normalizing the respective gene expression counts relative to threshold values for each gene;

selecting one or more member-comparison factors;

identifying one or more meta-genes relevant to the one or more member-comparison factors;

comparing normalized gene expression counts for the one or more meta-genes for the first member against normalized gene expression counts for the one or more meta-genes for the second and third members to generate a second dataset representing each member'"'"'s relative over-expression relative to each meta-gene,

receiving a selection of the first member from a user;

identifying a first angular assignment for the second member and a second angular assignment for the third member; and

displaying, via a display operatively connected to the processing device, the first member at a focal point of a graphical user interface on a display, the second member at a first radial distance based on the measure of similarity between the first and second member and the first angular assignment, and the third member at a second radial distance based on the measure of similarity between the first and third member and the second angular assignment.

17. ) The method of claim 16, wherein one of the one or more member-comparison factors is a specific type of disease state, the method further comprising:

receiving, via the graphical user interface, a user indication of the specific type of disease state from a user; and

selecting the meta-genes based on the user indicated disease state.

18. ) The method of claim 16, wherein the one or more member-comparison factors is a mutation of one or more genes in a genetic pathway, the method further comprising:

receiving, via the graphical user interface, a user indication of a specific gene from a user;

identifying one or more genetic pathways including the specific gene;

receiving, via the graphical user interface, a user indication of one of the one or more genetic pathways; and

selecting the meta-genes based on the user indicated genetic pathway.

19. ) The method of claim 16, wherein the one or more member-comparison factors is a mutation of one or more genes in a genetic pathway, the method further comprising:

receiving, via the graphical user interface, a user indication of one of the one or more genetic pathways from among a plurality of genetic pathways; and

selecting the meta-genes based on the user indicated genetic pathway.