= P-body detection threshold
{c}
{title2=$\tau$}
A \b[P-body detection threshold] is a decision rule used to classify candidate image features as detected <P-bodies>. It can involve intensity, contrast or size after <image segmentation>. The threshold belongs to the measurement and analysis procedure; it is not automatically a physical concentration at which <P-body assembly> begins.
In <fluorescence microscopy>, a labeled <protein> produces a signal that must be distinguished from background and other image features. Exposure, label abundance, noise and <background subtraction> affect visibility. A2005 study connected changes in visible P-bodies with RNA integrity and translation state, providing biological perturbations against which such image readouts can be interpreted.<Processing bodies require RNA for assembly and contain nontranslating mRNAs by Daniela Teixeira and others>[Abstract and Figure6, cellular foci and RNA-dependent assembly experiments]{ref}
For an original fixed-image example, suppose four candidate foci have measured scores $I=(2,5,5,9)$. Define a score threshold $\tau$ and count
$$
N(\tau)=\sum_{i=1}^{4}\mathbf1\{I_i\geq\tau\}.
$$
Then $N(5)=3$, while $N(6)=1$. The same four candidates and the same biological sample yield different reported counts because the rule changed. More generally, increasing $\tau$ cannot increase this count: every candidate passing a higher threshold also passes a lower one.
A <negative control> helps test whether the rule also selects background features. For example, if a stipulated control has scores $(2,3,6)$, the threshold5 still selects one candidate. That observation calls for interpreting specificity; it does not identify that control feature as a P-body.
A reported increase or decrease in detected foci should therefore state the scoring and detection conditions and be compared with suitable controls. Intensity alone does not establish molecular composition or <RNA decay> activity, and a focus falling below threshold need not mean that all of its molecules disappeared.
Back to article page
