01Ask whether observation changes the cells

Laissue and colleagues argue for quantitative, experiment-specific assessment of phototoxicity in live fluorescence imaging. Their reporting guidance recognises that one universal standard cannot cover the variety of specimens and biological questions.Laissue and colleagues — Assessing phototoxicity in live fluorescence imaging (opens in a new tab)

Alghamdi and colleagues experimentally examined blue-light exposure and motility in PC3 cells. Their study illustrates that an imaging-related perturbation can affect the behaviour being measured; its findings should not be converted into a universal illumination threshold for other models.Alghamdi and colleagues — Assessing Phototoxicity in a Mammalian Cell Line: How Low Levels of Blue Light Affect Motility in PC3 Cells (opens in a new tab)

For a peptide experiment, look for evidence that the imaging conditions preserve the outcome the authors want to study. A comparison should make clear whether the reference and peptide conditions experience equivalent observation.

The useful question is not simply whether the images look attractive. It is whether the study can distinguish a response to the experimental condition from a response introduced or modified by observing it.

02Understand what the frame schedule can resolve

Consider an original fictional sequence captured at minutes zero, five and ten. A visible change is absent at minute five and present at minute ten. Those observations place its detected onset within an interval; they do not identify the exact second it began.

This interval assumes that the feature was detectable when present and that its definition remained consistent. If the phenomenon can appear and disappear between frames, even more of its history may remain unobserved.

Likewise, movie playback speed is not biological speed. Read the elapsed-time label or acquisition interval before describing how quickly a cell moves or a response unfolds.

03Follow which cells remain in the analysis

An original fictional movie begins with 50 trackable cells, but only 30 complete tracks appear in the final analysis. The reader needs to know what happened to the remaining 20 before treating the displayed average as a summary of all starting cells.

Some may leave the field, become difficult to distinguish or fail an analysis rule. Those possibilities are not interchangeable with a biological nonresponse. The study should explain the exclusion categories rather than leaving their meaning implicit.

Inspect whether tracks were selected before or after the response was visible. Selection based on especially clear or striking responses can answer a different question from analysis of a consistently defined starting population.

When cells cross paths, ask how identities are maintained and how uncertain links are handled. A smooth trajectory produced by software is still an interpretation of successive observations, not independent proof that every identity assignment is correct.

04Combine the movie with a defined measurement

A representative movie can explain the appearance of a response, while quantified observations establish how often or how strongly it occurred in the analysed material. Read both roles separately. One striking sequence cannot establish a population frequency.

For a peptide study, state what was tracked: position, shape, a fluorescent marker or another feature. If the authors infer a mechanism from that feature, identify the additional evidence connecting the two.

Preserve the number of independent experiments as well as the number of cells or frames. Many images of the same culture provide temporal detail, but do not by themselves describe variation between independently prepared cultures.

A strong evidence note includes the observation interval, analysis population and assessment of imaging effects. It then states the measured response and the main remaining uncertainty, without treating a compelling movie as a complete explanation of the biology.