The single most time consuming and expensive part of a population PK analysis is the model selection. At Nuventra we are changing the paradigm for delivering consulting services that provide useful information at a reasonable cost and in a time frame consistent with decision making in the real world. DARWIN is a software program that automates the selection of a population PK model. DARWIN is based on the computational mathematics of survival of the fittest and evolution.
DARWIN is the next generation in population PK modeling with results that are:
Dr. Mark Sale has created a proprietary method for automated model selection in population pharmacokinetic (Pop PK) analyses that SAVES TIME AND MONEY. This approach uses a method that is well established in the engineering world for finding optimal solutions called “Genetic Algorithm” or “GA”.
Genetic Algorithm is a global search algorithm. That is, the user specifies all the things that should be considered in looking for the optimal solution. In the engineering world, GA has been used for a remarkable variety of difficult problems by looking for the optimal solution among many different combinations of options. GA has been used to find optimal designs for integrated circuits, aircraft, and manufacturing systems where there are just too many options to evaluate but all of which must be just right for the system to work optimally. In the case of pharmacokinetic models, the different options include the number of compartments, the absorption model, the random effects (both residual and between subject), and the covariates (such as demographics, hepatic & renal dysfunction, concomitant medications, etc.). The algorithm then searches for the optimal combination of these features, looking for the best model, by objective measures, to use in population PK analyses.
The Genetic Algorithm is based on the mathematics of survival of the fittest and evolution. In general biology, the algorithm results in the continuous improvement in the combination of genes (Darwin’s classic theory of evolution). In the case of the Genetic Algorithm for pharmacokinetic drug development model selection, the algorithm results in the continuous improvement in the combination of the pharmacokinetic features (number of compartments, absorption model, covariates, etc.), ultimately finding the optimal combination of those features, and the “best” model among all the possible combinations of features.
- Unsupervised machine learning-based mathematical model selection US patent # 7,085,690
- “Application of a single-objective, hybrid genetic algorithm approach to pharmacokinetic model building” J Pharmacokinet Pharmacodyn (2012) 39:393–414.
Advantages of the Pharmacokinetic Genetic Algorithm:
- Robust population PK model selection
- Faster population PK analyses vs. traditional methods
- Cost savings for population PK analyses due to computational efficiency in implementing the genetic algorithm
- Usable information to make informed decisions in drug development
- Shorter timeframes to critical go, no-go decisions based on data