You’ve run your western blot and you’re staring at the developed film or imager output. The signal is there, but is it real? Is the size right? What about that extra band? Reading a western well takes practice, but the framework is straightforward.
What you’re actually looking at
A western blot displays proteins separated by molecular weight, transferred to a membrane, and detected with a target-specific antibody. Each band represents protein at a specific size range that bound your antibody.
Step 1: Check the molecular weight
- Expected size: The predicted molecular weight from UniProt or a published reference
- Actual migration: Some proteins run smaller or larger than expected due to glycosylation, phosphorylation, charge, or unusual amino acid composition
- Confidence interval: ±10% of expected size is normal; ±20% suggests modification or fragmentation
Step 2: Evaluate the loading control
- Even bands across lanes — equal loading
- Variable bands — uneven loading; your “differential expression” might be a loading artifact
- Total protein normalization (Ponceau or stain-free) is more rigorous than housekeeping proteins
Step 3: Quantify the signal
Don’t trust your eyes. Use ImageJ or commercial imaging software:
- Open the image in ImageJ
- Use the rectangular selection tool to draw boxes of equal size around each band and around a background region
- Measure mean intensity (Analyze → Measure)
- Subtract background from each lane
- Normalize to loading control
Avoid saturated bands — quantification is only valid in the linear range.
Step 4: Interpret common band patterns
Multiple bands
- Isoforms: Many proteins exist as multiple splice variants
- Post-translational modifications: Phosphorylation, glycosylation, ubiquitination shift apparent size
- Proteolytic fragments: Degradation during sample prep produces smaller fragments
- Cross-reactivity: Validate with knockout/knockdown
No band
- Insufficient protein
- Antibody dilution too high
- Transfer failure (check Ponceau)
- Protein degraded during sample prep
- Wrong cell type or condition
Smeared bands
- Sample degradation
- Overloading
- Buffer issues during electrophoresis
Unexpectedly larger band
- Post-translational modification (phospho-protein, glycosylation, ubiquitination ladder)
- Dimer or oligomer that didn’t denature fully
- Fusion protein or tagged construct
Unexpectedly smaller band
- Proteolytic fragment
- Truncated splice form
- Non-specific antibody binding
Step 5: Verify with controls
- Positive control: a sample known to express the target
- Negative control: a knockout or knockdown of the target
- No-primary control: rules out direct binding by the secondary antibody
- Loading control: verifies even sample loading
Reporting standards
Many journals now require: full uncropped blot images in supplementary data, molecular weight ladder visible in every blot, quantification across at least three biological replicates, antibody catalog numbers and lot numbers, statistical analysis with appropriate tests.
A clean western blot is a story with multiple chapters: size, intensity, controls, and replication. Interpret each layer, and don’t over-conclude from a single image.
