Information processing method for evaluating biochemical pathway models using clinical data

• US 20040167763A1
• Filed: 11/20/2003
• Published: 08/26/2004
• Est. Priority Date: 05/31/2002
• Status: Abandoned Application

• Current Assignee: Prosanos Corp. (The Cigna Group)
• Original Assignee: Prosanos Corp. (The Cigna Group)
• Inventors: Liebman, Michael N
• Application Number:
  US10/478,313
• Publication Number:
  US 20040167763A1
• Time in Patent Office: Days
• US Class Current: 703/11
• CPC Class Codes: G16B 5/00 ICT specially adapted for modelling or simulations in systems biology, e.g. gene-regulatory networks, protein interaction networks or metabolic networks
• CPC Class Codes: G16B 5/20 Probabilistic models
• CPC Class Codes: G16B 5/30 Dynamic-time models

Abstract

A method of evaluating a biochemical pathways models using clinical data, includes representing the biological model within a computing system, in the form of a hierarchy of Petri nets or stochastic activity nets, wherein the nodes of the net represent biological or biochemical components, and the arcs correspond to the flow of biochemical components in the model. A time series of measurements of some of the biochemical components described by the model are recorded and an observed pattern of relationships between inputs and outputs of the respective nodes is compared to an expected pattern of the relationships to determine whether the model describes the behavior of the biochemical components.

Claims

1. A method of evaluating a biochemical pathways models using clinical data, comprising the steps of:

(a) representing the biological model within a computing system, in the form of a hierarchy of Petri nets or stochastic activity nets, wherein the nodes of the net represent biological or biochemical components, and the arcs correspond to the flow of biochemical components in the model;

(b) recording a time series of measurements of some of the biochemical components described by the model, wherein these measurements are made on samples from one or more humans or other living organisms;

(c) storing those observations as a data set in the computing system, which performs the subsequent steps as automated computations;

(d) for each time series in the stored data set, calculating the rate of change of some or all variables with respect to time, and augmenting the data set with that information;

(e) for each input to a network node, labeling the input with the expected effect of an increase in token flow at the input on each output of the node, either “

increasing”

or “

decreasing”

;

(f) comparing an observed pattern of increases and decreases of measured biochemical levels versus the expected pattern of increases and decreases in flow rates of tokens through the corresponding nodes; and

(h) marking the biochemical components for which the expected effect is an increase while the biochemical measurements show a decrease, or vice versa, wherein the marking indicates that the model does not describe the behavior of those biochemical components.

2. The method according to claim 1, wherein step (d) comprises selecting a subset of the data set based on patient demographic data or a history of prior treatment before calculating a rate of change of some or all variables from the selected subset of the data set.

3. The method according to claim 1, further comprising:

for models that have no markings in step (h), identifying nodes whose outputs are connected to inputs of a number of nodes that is higher than the average connectivity of outputs of nodes to inputs of other nodes.

4. The method according to claim 3, wherein the identified nodes are identified as key regulatory factors.

5. A method of evaluating a biological model comprising the steps of:

(a) representing the biological model within a computing system, in the form of a hierarchy of Petri nets or stochastic activity nets, wherein the nodes of the net represent biological or biochemical components, and the arcs correspond to the flow of biochemical components in the model;

(b) recording a set of measurements of some of the biochemical components described by the model, wherein these measurements are made on samples from one or more humans or other living organism patients, wherein the set of patients is divided into subsets representing patients with a disease, various diseases, and/or healthy patients;

(c) storing the measurements a data set in the computing system, which performs the subsequent steps as automated computations;

(d) for subsets of the data, calculating the change of some or all measured variables with respect to the corresponding value of that variable for healthy patients or diseased patients in a different subsets, and augmenting the data set with that information;

(e) for each input to a network node, labeling the input with the expected effect of an increase in token flow at the input on each output of the node, either “

increasing”

or “

decreasing”

;

(f) comparing the pattern of increases and decreases of measured biochemical levels in various disease states versus the expected pattern of increases and decreases in flow rates of tokens through the corresponding nodes; and

(g) marking the biochemical components for which the expected effect is an increase while the biochemical measurements show a decrease, or vice versa, the marking indicating that the model does not adequately describe the behavior of those biochemical components.

6. The method according to claim 5, further comprising:

(h) for models which are consistent in that they have no markings from step (g) above, further marking nodes whose outputs are connected, directly or indirectly, to inputs of a relatively large number of nodes compared to the average connectivity of the nodes in the network, the marking indicating that the node is a potential key regulatory factor.

7. The method according to claim 5, wherein step (d) comprises further restricting the subsets of the data set, based on patient demographic information or history of prior treatment;

8. The method according to claim 6, wherein the step (b) comprises recording measurements of a subset of patients comprising only diseased patients;

and step (h) comprises using cluster analysis to stratify the disease into clusters.

9. The method according to claim 6, wherein the step (b) comprises recording measurements of both diseased and healthy patients, and step (h) comprises using the identified nodes for developing a biochemical test for the disease being studied.

10. The method according to claim 6, wherein the step (b) comprises recording measurement of patients in various stages of a disease, and step (h) comprises using the identified nodes to measure or predict disease progression stages.