Optimal experiment design in plasma protein metabolic studies: Sequential optimal sampling schedules for quantifying kinetics

Abstract

It is not difficult to demonstrate, even to the novice investigator, that methodological problems in the clinical/physiology/biochemistry laboratory environment tend to have a significant limiting effect on the types and complexity of mathematical models of biological regulatory systems that can be constructed appropriately. This is especially true when we deal with quantitative questions about systems that function with molecular signals, such as metabolic, endocrine and pharmacokinetic systems. Naturally, this restricts the kinds of questions that can be addressed, as well as the accuracy of results that can be obtained with the aid of such models1. It is common knowledge among life science investigators that a very limited number of ‘access ports’ for introducing test inputs and measuring output responses are normally available in the in vivo state—especially in human studies, typically blood and urine. Also, the number of samples that can be obtained and the accuracy at which these samples can be assayed are, more often than not, severely limited by practical constraints. These include blood volume depletion problems and subject availability scheduling. This is in stark contrast to the far more favourable environment for data collection and information retrieval in most technological, or non-living, systems.