Wild-type and mutant predictive value.

Case 1 refers to the situation whereby all events, regardless of event lifetime, are analyzed. The simulated 50∶50 mix dataset forms the individual events to be protein-called by which wild-type and mutant predictive value is determined. These events are randomly selected from the total number of wild-type and mutant events. After selection (and removal from the total number of events), the remaining events form the training sets for both wild-type and mutant. Wild-type predictive value refers to the likelihood that an event is in fact a wild-type event given that the protein-calling algorithm has called it as wild-type. Similarly the mutant predictive value refers to the likelihood that an event is in fact a mutant event given that the protein-calling algorithm has called it as mutant. Case 2 refers to the case whereby only those events that have an event lifetime of ≥1 s are analyzed (i.e. only long time events makeup the training sets for both wild-type and mutant and the generated 50∶50 mix dataset). The predictive value, not surprisingly, improves when only considering long-time events which is primarily due to the fact that long-time events can be better assessed in terms of their kinetics than short-time events (i.e. the amount of data available to characterize an event is proportional to the event lifetime) thereby improving protein-calling accuracy.