False positive fraction

One measure of accuracy is the likelihood ratio (LR) [17]. The likelihood ratio is based on the relationship between the true positive fraction and the false positive fraction. Thus, LR = TPF/FPF. Substituting terms, the LR = sensitivity/(1-specificity). Receiver-operating characteristic curves are graphical summaries of likeli-... 

To provide a method for the evaluation, or at least comparison between two analyzers probabilities of false alarms, a model for the prediction of FAR was devised, which is based on relating the total number of the analyzer s orthogonal measurement channels and the analyzer s signal-to-noise ratio (R, ) tio to the probability of analysis errors obtained imder specific test conditions, which also corresponds to errors predicted by ROC curves [8], as illustrated by the example in Fig. 9.3.14. The area above the ROC curve, 1-Az, represents the total instrumental error of the involved analyzer, and may be plotted by simply using the result of tests yielding values of FPF (False Positive Fraction) and FNF [16]. One set of specific test conditions assumed in early model computations consisted of a simple symmetry for FPF and FNF, leading to the symmetrical Decision Matrix and Proportions shown in Fig. 9.3.14. 

Table 9.3.1 Estimated, relative false positive fraction (EPF) and area above the ROC curve for analyzer examples.

Where A tp is the number of true positives, A/pw is the number of false negatives, A/jn is the number of true negatives, A pp is the number of false positives, and N = A/pp + AlpN + A/tn + A fp- Sensitivity and specificity are often expressed as either fractional quantities ranging from 0 to 1, or percentage quantities ranging up to 100%. In either case, 1.0 or 100% for both metrics indicates perfect method performance. 

To begin an examination of the fraction of false positives, we chose 14 of our cycling mRNAs at random and examined their cycling by real-time PCR. We could clearly confirm the cycling and microarray patterns for 10 of the 14 mRNAs, and the cycling was likely positive for two more. Only in two of the 14 cases was circadian cycling unlikely, based on the real-time results. We conclude that most of our 134 mRNAs are real cyclers and that false positives constitute only a minority of the 134 mRNAs. The number 134 is probably a gross underestimate. 

Substitution of isobutane for a small fraction of the airstream increases the quenching of spurious HO fluorescence in the background channel by about 1%, so exact cancellation is not achieved in subtraction and a false positive signal is introduced (80, 103, 107). This situation applies not only to ozone photolysis but also to all other photolytic HO sources. 

The general aim of VS methods is to retrieve from a molecular database a fraction of true positives that is significantly larger than that of a random compound selection. If a VS method selects n molecules from a database with N entries, the selected hit list consists of active compounds (true positive compounds, TP) and decoys (false positive compounds, FP). Active molecules that are not retrieved by the VS method are defined false negatives FN), whereas the unselected database decoys represent the true negatives TN) (Figure 3.3). 

---