Browse the research registry public · Customer Self-Service

Research registry ID

Date submitted

Project lead

Title

AI-assisted genotype-phenotype relation prediction using genome-scale metabolic models

Community 1

Community 2

Community 3

Lay summary

Metabolism is a process by which all cells manufacture critical products they need to grow and divide. Cancer cells have altered metabolism to healthy cells, which allows them to grow and divide much faster. Standigm has constructed a genome-scale model(GSM)-based approach to systematically investigate human metabolism. GSM consists of a comprehensive network of human metabolic reactions and enables us to reproduce metabolic networks under diverse biological conditions. The GSM maps out a comprehensive network of biochemical reactions that happen in human cells. This allows us to model the different biochemical reactions we can expect to find happening within a cell, based on its genome, to help predict likely levels of these metabolites.
The strength of metabolic reactions is quantified as flux values (relative amount of metabolic products made within a specific time frame). Given that any biological system aims to increase its fitness, we can simulate how the reactions should be balanced to achieve the goal and thus predict the flux values for all reactions. This is called flux balance analysis(FBA). Based on this, Standigm can construct models that link the flux values to the genome sequence and clinical data. The model can predict how a certain genetic variant affects the phenotype (clinical expression of the trait). Through this approach, we can explain how certain genetic variants result in altered reactions which in turn contribute to the phenotype. This can be applied to find novel anticancer targets and mechanisms of action to benefit drug discovery.

In this section