How to Properly Label a Western Blot for Publication

A properly labeled Western blot figure includes molecular weight markers, clear lane annotations, identified antibodies, visible loading controls, and enough surrounding context that another scientist could evaluate the result without guessing. Yet a large-scale review of published papers found that more than 95% of Western blot figures lacked molecular weight markers entirely, and roughly a third had no molecular weight labels of any kind.¹ Getting the labeling right is not just cosmetic; reviewers increasingly reject figures that omit key information, and journals are tightening their requirements for raw blot data alongside the finished figure.

What Every Western Blot Figure Needs on It

Think of the figure itself as the first layer of labeling. The image that appears in your manuscript should let a reader identify, at a glance, what protein is being detected, which samples are in which lanes, and where on the molecular weight scale the band falls. At minimum, the blot image should include:

Molecular weight markers: Show the positions of at least two size standards flanking your band of interest. If your protein runs at about 45 kDa, include markers at roughly 35 and 55 kDa (or whatever your ladder provides). Mark these with their kilodalton values directly on or beside the image.
Lane labels: Every lane needs an identifier. For a treatment experiment, that means labeling each lane with the condition, dose, or time point. For tissue or cell-line comparisons, name the source. Biological replicates should be distinguishable from technical replicates.
Target protein identity: State which protein the antibody detects, either above or beside the blot panel. If you are showing a phosphorylated form, specify the phosphorylation site.
Loading control: The corresponding loading control blot, labeled with the housekeeping protein name or “Total protein,” should appear directly beneath or beside the target blot so readers can visually compare lane-to-lane loading.

The absence of molecular weight information is one of the most common problems in published blots. When a study surveyed hundreds of papers across neuroscience and cell biology, over 90% presented only cropped blots, and the vast majority had no visible molecular weight markers in any blot image.¹ Without markers, there is no way for a reader to confirm that the detected band is the expected size, which undercuts the entire point of the experiment.

The Figure Legend Does Heavy Lifting

The blot image carries the visual labels, but the figure legend carries the experimental context. Journals differ in exactly how they want this formatted, but the information that should appear in every Western blot legend includes the antibody supplier and catalog number (or Research Resource Identifier, known as RRID), the species the antibody was raised in, the dilution used, the detection method, and how long the blot was exposed. If you stripped and reprobed the membrane, state that clearly, including the order of probing.

Methods sections in published Western blots frequently omit these details. The same large review that flagged missing markers also found that antibody identifiers like catalog numbers and RRIDs were regularly left out, especially for secondary antibodies, and that blocking conditions and protein loading amounts were often absent from the methods text.² These omissions make it functionally impossible for another lab to reproduce the result, which is the practical definition of insufficient reporting.

A useful habit is to write the figure legend before you finalize the image. If you cannot fill in the antibody catalog number, the protein amount loaded per lane, or the exposure time while drafting the legend, that is a sign you did not record it during the experiment and need to track it down before submission.

Choosing and Presenting a Loading Control

Loading controls show that roughly equal amounts of protein were loaded in each lane. For years, researchers defaulted to probing for a housekeeping protein like beta-actin, tubulin, or GAPDH. These are still widely used, but the approach has real weaknesses. Housekeeping proteins can vary with experimental treatments, saturate the membrane at typical loading amounts, and have a narrow dynamic range that makes them unreliable for detecting differences in loading above small quantities.³

Total protein staining has emerged as a stronger alternative. Methods like Ponceau S, Coomassie Brilliant Blue, and stain-free gel technology capture the full protein load in each lane rather than relying on a single housekeeping gene product. Ponceau S staining, for instance, has been shown to balance accuracy and precision better than actin or GAPDH for normalizing loading variability.⁴ Stain-free technology, which uses trihalo compounds built into the gel to image total protein without any post-transfer staining step, performed as well as or better than Ponceau S and outperformed beta-actin.⁵

When you present a loading control in the figure, label it clearly. If you used total protein staining, include an image of the stained membrane and label it “Total protein” or with the specific stain name. If you used a housekeeping protein, label the panel with the protein name. Either way, the loading control should appear in the same figure panel as the target blot, not buried in supplemental data where readers might miss it.

Cropping, Splicing, and How Much of the Blot to Show

Cropping a blot to show only the region around the band of interest is standard practice and accepted by journals, but it introduces problems if done carelessly. A tightly cropped band removes all context about whether nonspecific bands were present, whether the molecular weight is correct, and whether the background was clean. Over 90% of blot images in published neuroscience and cell biology papers are cropped.¹ The crop itself is not the issue; the issue is that most cropped blots lack the flanking molecular weight markers needed to confirm band identity.

If you crop, keep enough of the image to include at least one molecular weight marker on either side of the band. Indicate clearly if lanes have been rearranged or if non-adjacent lanes have been placed side by side. Spliced images should be marked with a thin vertical line or gap between non-contiguous lanes, and the figure legend should state explicitly that lanes were rearranged. Some journals now require that any splice be labeled on the figure itself, not just in the legend, because readers often skip legends.

Many journals now ask for full-length, uncropped blot images as supplementary material or as part of peer review submission, even if the published figure uses a cropped version. Making this source data available lets editors and reviewers check for problems that cropping might conceal, such as duplicated bands or unexpected nonspecific binding.⁶ Get in the habit of saving uncropped images immediately after acquisition, before any editing, in the original file format.

Quantification and How to Present It

Western blots were once treated as purely qualitative, showing that a protein was present or absent. That era is largely over. Reviewers and editors now routinely expect quantitative interpretation, meaning you need to show fold changes in protein expression between conditions, supported by densitometry and statistical analysis.⁷

The standard approach is ratiometric analysis. You measure the signal intensity of your target band, divide it by the corresponding loading control signal for the same lane to get a normalized value, and then express each sample’s normalized value as a fold change relative to a control sample. A fold change above 1.0 means the protein is more abundant than in the control; below 1.0 means less abundant.⁸ These fold change values are unitless and proportional, which makes them comparable across experiments as long as the normalization method is consistent.

Present quantification data as a bar graph or dot plot alongside the blot image, with individual data points shown and error bars representing the standard deviation or standard error of the mean. State in the legend how many biological replicates are represented, which statistical test was used, and what your threshold for significance was. Reviewers will push back on quantification claims supported by a single blot or by blots without the underlying statistical analysis.

One thing to watch: quantification is only meaningful within the linear range of your detection system. If the signal is saturated, the densitometry numbers are meaningless regardless of how carefully you normalize them. This is where your detection method matters. Fluorescent secondary antibodies provide a broader quantifiable range than chemiluminescence, roughly tenfold greater, though with somewhat lower sensitivity.⁹ If you are making quantitative claims, fluorescent detection is generally the better choice. State the detection method in your figure legend either way.

Antibody Validation and Controls Worth Showing

A labeled blot is only as trustworthy as the antibody that produced the signal. The gold standard for proving antibody specificity is a knockout control: you run a sample from cells or tissue where the target gene has been deleted. If the band disappears in the knockout lane, the antibody is genuinely detecting your protein of interest. If the band persists, that is evidence of off-target binding, which means your beautifully labeled figure may be reporting the wrong protein.¹⁰

Not every experiment can include a knockout control, but you should always include some form of positive and negative control. A positive control is a sample you know expresses the target protein. A negative control is a sample that should not express it, or a lane where the primary antibody was omitted. Including purified protein as a positive control, when available, gives readers confidence that the band at the expected molecular weight is the real target.¹¹

In the figure, label control lanes clearly: “KO,” “WT,” “No primary Ab,” or whatever applies. In the legend, explain what each control lane contains. If the antibody has been validated in a previous publication, cite that paper. If you validated it yourself using knockout or knockdown experiments, consider including the validation blot as a supplementary figure.

Multiplex Blots and Phosphorylation Panels

Fluorescent detection makes it possible to probe two or more targets on the same membrane simultaneously, using secondary antibodies labeled with spectrally distinct fluorophores. This is especially useful when you need to compare a phosphorylated protein to its total form, because both proteins run at the same molecular weight and cannot be separated by cutting the membrane. Simultaneous detection with fluorescent secondaries from different host species avoids the need to strip and reprobe, which can cause signal loss and incomplete stripping artifacts.¹²

When labeling a multiplex blot, each channel needs its own panel. Show the merged image if it helps the reader see co-localization, but always include the individual channels as separate panels so the reader can evaluate each signal independently. Label each panel with the target protein and, ideally, the fluorophore color or wavelength used. In the legend, specify which primary antibody was raised in which species and which secondary antibody (and fluorophore) was used for each channel. If you combined a rabbit anti-phospho antibody with a mouse anti-total antibody, state that pairing explicitly.

For phosphorylation studies, the figure typically shows the phospho-protein blot, the total protein blot, and the loading control. The quantification graph then plots the ratio of phospho signal to total protein signal, normalized to the loading control. Label the y-axis of the graph clearly, something like “p-Protein / Total Protein (fold change).” Ambiguous axis labels are a common reason reviewers request figure revisions.

Image Adjustment and What Crosses the Line

Adjusting brightness and contrast is allowed and often necessary for a clear figure. The rule is that any adjustment must be applied uniformly across the entire blot image, not selectively to individual lanes or bands. You cannot darken one band while leaving others unchanged, and you cannot erase background from a region around your band of interest while leaving the rest untouched. Nonlinear adjustments like gamma correction are generally discouraged for quantitative blots because they distort the relationship between signal intensity and protein abundance.

State in the figure legend or methods section whether any image processing was applied and what software was used. If you adjusted brightness and contrast, say so. If you used ImageJ, Fiji, LI-COR Image Studio, or Bio-Rad Image Lab for densitometry, name the software and version. This transparency is not optional in most journals’ current guidelines and protects you from accusations of image manipulation during review or after publication.

Save your original, unprocessed image files in the acquisition software’s native format before making any adjustments. TIFF files are preferred for publication because they are lossless. Never save a blot image as a JPEG for submission; the compression introduces artifacts that can alter band appearance and make post-publication forensic analysis unreliable.

Archiving Raw Data and Supplementary Files

Beyond what appears in the figure, journals increasingly want the underlying raw data to be accessible. This includes the uncropped blot images, any images of replicate blots not shown in the main figure, total protein stain images, and the raw densitometry data. The argument for publishing source data is straightforward: it lets editors, reviewers, and readers verify the result and check for errors or manipulation.⁶

In practice, most published papers do not include full blot source data in the supplement. The same review that identified widespread marker omissions found that source data were generally unavailable to readers.¹³ This is changing as more journals adopt mandatory data availability policies, but the gap between policy and practice is still wide. If you are preparing a submission, check whether the target journal requires uncropped blots at submission, at revision, or as published supplementary material. Some require all three.

When you assemble supplementary files, label them systematically. Each supplementary blot image should be named or captioned to match the main figure panel it supports. Include annotations on the uncropped image indicating which portion was used for the published figure, using a box or bracket overlay. This saves the reviewer from having to guess which region corresponds to Figure 3B.

A Practical Checklist for Formatting the Figure

Before you submit, run through the figure with fresh eyes and check for these items. Catching these in your own review is far better than catching them in a revision letter.

Molecular weight markers: Visible on the blot image, not just listed in the legend. At least one above and one below the band of interest, labeled in kDa.
Lane labels: Every lane identified with sample name, condition, or replicate number. No unlabeled lanes.
Target protein name: Stated on or directly adjacent to each blot panel.
Loading control panel: Shown in the same figure, aligned with the target blot so lane correspondence is obvious.
Quantification graph: If claiming changes in expression, include a graph with individual data points, error bars, sample sizes, and the statistical test used.
Splice indicators: Any rearranged or non-contiguous lanes marked with a visible dividing line on the image.
Legend content: Antibody names, catalog numbers or RRIDs, host species, dilutions, secondary antibody details, detection method, exposure time, software used for image processing and densitometry, number of biological replicates.
Image format: TIFF or high-resolution PDF, not JPEG.
Supplementary files: Uncropped blot images annotated to match main figure panels, raw densitometry data, replicate blots if available.

Missing any one of these items is unlikely to sink a paper, but missing several will almost certainly trigger reviewer comments and slow down the review process. The researchers who have the smoothest time at review are usually the ones who treat figure preparation as part of the experiment rather than an afterthought.

When Journals Disagree on Requirements

There is no single universal standard for Western blot figure formatting. Some journals require uncropped blots at initial submission; others only ask for them if a reviewer raises a concern. Some insist on total protein staining as a loading control; others still accept housekeeping proteins without complaint. Some want RRIDs for every antibody; others accept catalog numbers alone. The variability is real and can be frustrating when you are reformatting a rejected manuscript for a new journal.

A few resources can help you navigate this. The journal’s author guidelines usually have a section on figure preparation and image integrity, and many journals now publish specific Western blot checklists. If the guidelines are silent on a particular detail, aim for the most stringent standard you are aware of. Including molecular weight markers, full antibody details, total protein stain loading controls, and uncropped supplementary images will satisfy virtually any journal’s requirements, even if not all of those items are formally mandated. Reviewers notice when a figure is thorough, and it builds trust in the data before the science is even evaluated.