System and method for identifying networks of ternary relationships in complex data systems
View Patent ↗A system and method for identifying high order associations between variables in complex systems that is particularly useful where there is no correlation or weak correlation between variables due to the influence of a third variable, a ternary relationship. The ternary relationship describes how the variation in the pattern of association between a pair of variables, including its sign and strength, is mediated by a third variable. In one embodiment applied to gene expression data, the activity of pairs of correlated genes due to the activity of one or more third genes is shown.
1. A method for identifying patterns of gene expression within a cell of an organism, comprising:
sequencing the genome of a cell to identify genes;
measuring the expression of said genes over time;
correlating the activity of each gene with the activity of every other gene in the genome to produce pairs of correlated genes;
associating the activity of each gene with the correlated activity of said pairs of correlated genes;
identifying scouting genes from the associated activity of each gene with the activity of said pairs of correlated genes; and
evaluating the biological characteristics of the products of said scouting genes and said pairs of correlated genes.
2. A method as recited in claim 1 , further comprising:
identifying the association of a first scouting gene with a second scouting gene to identify patterns of gene expression and suppression in an organism.
3. A method as recited in claim 1 , wherein said measuring step comprises:
measuring a quantity of mRNA for each gene over time as a measure of gene expression.
4. A method as recited in claim 1 , further comprising:
investigating a biological activity of said identified scouting genes in a cell.
5. A method as recited in claim 1 , further comprising:
modeling biological activity of said identified genes in the organism.
6. A method for identifying patterns of irregular gene expression in tumor cells, comprising:
measuring gene expression for a plurality of genes in a normal cell;
correlating the activity of each gene with the activity of every other gene of said plurality of genes to produce pairs of correlated genes;
associating the activity of each gene with the correlated activity of said pairs of correlated genes to form a baseline pattern of normal gene activity;
measuring gene expression for a plurality of genes in a tumor cell;
correlating the activity of each gene with the activity of every other gene of said plurality of genes to produce pairs of correlated tumor cell genes;
associating the activity of each gene with the correlated activity of said pairs of correlated tumor genes to form a baseline pattern of tumor gene activity; and
comparing the baseline pattern of normal gene activity with the baseline pattern of tumor gene activity to determine differences in the activity patterns.
7. A method as recited in claim 6 , further comprising:
evaluating the biological properties of the gene products of the genes that are expressed differently in the tumor cell than in the normal cell.
8. A method as recited in claim 6 , wherein said associating step of said normal cell measurements further comprises:
identifying scouting expression genes from the associated activity of each gene with the activity of said pairs of correlated genes; and
evaluating the biological characteristics of the products of said scouting genes and said pairs of correlated genes.