**5. Method-dependent biological quantities**

Some important biological quantities are not described in the SI unit of mole. Quantities describing activity, viability, or potency used in biology communities are defined by the methods of assessment that are agreed among the experts in the related fields. Comparability of measurements of these quantities is governed by how precisely the measurement protocols are executed. Therefore, metrology for such quantities needs to be established to support 'metrological execution' of the given measurement protocols. In this regard, improvement of measurement comparability for these quantities can be achieved by somewhat more than providing a simple reference material or reference measurement procedure. In a typical measurement protocol of an enzyme activity, for example, several experimental conditions such as the quantity of the substrate, reaction temperature, pH and buffer composition influence the measurement results. Comparability of measurement of enzyme activity can be assured only when such experimental conditions are tightly controlled. In other words, measurement uncertainties of such quantities arise from the uncertainties in the control of such reaction conditions. It needs to be stressed that such experimental conditions are metrological quantities that can be precisely measured and controlled with clear uncertainty budgeting. Application of metrological rigorousness to such experimental conditions can dramatically improve the measurement comparability for method-dependent quantities. Although yet to be demonstrated, the use of modern computing power will also become essential in dealing with such complex measurement procedures affected by multiple variables. Highly confusing measurement uncertainties of such procedures may be only correctly budgeted by correct computer- simulations of the procedures.

The potential impact of such notion for method-dependent quantities has been demonstrated in polymerase chain reaction (PCR) (Yang et al., 2005). As described in earlier section, PCR is frequently used as a way to quantify genetic materials, especially in the format of real-time PCR (or q-PCR). In a PCR process, there are several experimental conditions that need to be tightly controlled. Among them, reaction temperature can be a critical control parameter as melting, annealing, and elongation steps of PCR are all profoundly affected by temperature. Numerous commercial thermal cyclers for PCR are available. How well do those thermal cyclers operate if we observe them with a metrological criticism in terms of temperature control? In our investigation (Kim et al., 2008), most PCR instruments showed satisfactory accuracy in static temperature control. However, in dynamic temperature control for PCR, substantially different performances were observed (Fig. 5.1).

Fig. 5.1. Real-time monitoring of in tube-temperature during PCR cycles for different commercial thermal cyclers. Unlike the ideal case (A), shortened annealing time due to slow response (B), or overshooting (C) and undershooting of temperature were observed (data from Bioanalysis Laboratory at KRISS).

Slower acting instruments did not reach the programmed temperature in a fast PCR. On the other hand, some fast acting instruments showed under- or overshooting, which could lead to drastically different results. Such imperfection of instruments is not well known, and proper measures against such imperfections are not well taken, which would cause substantial disparity in performing bioassays based on PCR. It will be very helpful if the degree of imperfection of instrumentation is analyzed, and is considered in preparation of a measurement protocol. For example, an extreme PCR speed needs to be avoided for users of slower-acting thermal cyclers.

Various bioassays of method-dependent quantity are crucial in QC or QA of biological products. Confidences on biological products are heavily dependent on the credibility and comparability of such bioassays for QC and QA. Therefore, the establishment of metrology for method-dependent quantities needs to be promptly established. Although brief, an example of how to achieve metrology for such quantities is described above.
