Introduction

Drug development is a lengthy and costly procedure with the majority of the tested compounds eventually rejected because they are deemed unsafe or not sufficiently able to lower disease risk. Modeling of the complex network of cause-and-effect relationships between biological molecules and disease will help us determine how changes in one factor affect all others. This will result in a system that can be used to accelerate drug development for the treatment of complex diseases such cardiovascular disease and type 2 diabetes.

Explaining the science

Mendelian randomisation (MR) is a well-established epidemiological method able to emulate a randomised control study and identify causal from confounded associations using genetic information. The method makes use of the genetic variation, randomised at conception, as a proxy measure (instrument) for the trait of interest. Using this, we can distinguish which pairs of metabolic traits are part of the same causal chain, from those correlated through other mechanisms. We found, however, that this approach could not distinguish precedence in the effect chain for traits situated in neighbouring links of this chain without additional evidence. In contrast, time-dependent measures of correlation are able to identify temporal precedence in a series of effects and, thus, provide the additional evidence required, though this does not always mean that a mechanistic causal relationships is present. Combining the inferred time-dependent causal associations with the MR approach can help address the disadvantages of each method by providing both a mechanistic view of the system and identifying the dynamic flow of information through it.

Project aims

Drug development is a lengthy and costly procedure with the majority of the tested compounds eventually rejected because they are deemed unsafe or not sufficiently able to lower disease risk. Modeling of the complex network of cause-and-effect relationships between biological molecules and disease will help us determine how changes in one factor affect all others. This will result in a system that can be used to accelerate drug development for the treatment of complex diseases such cardiovascular disease and type 2 diabetes.

Applications

The approach could be used to gain an early prediction of the safety and disease effect of pharmacological compounds in early stages of drug development. This will benefit the pharmaceutical industry in their effort to identify effective and safe medication and ultimately the patients suffering from complex diseases with a metabolic basis.
 

Recent updates

The project has not yet started but we do have the Mendelian randomisation relationships already available.

Funders

Collaborators