You bought an antibody, and it didn’t work, so you switched to another supplier. But that antibody hasn’t worked either! Arghhhh…
Writer Rita Mae Brown said, “Insanity is doing the same thing over and over again and expecting different results”. But is there something else going on here?
Before you blame your samples, your pipetting skills, or your bad luck, it’s worth noting that it could be the manufacturer’s fault.
This post is for anyone who buys reagents to help you understand what you’re actually purchasing when you pick a supplier, and a handful of checks that can catch a bad purchase before it costs you grant money.
Choose a free resource to help you move forward
REFERENCE CARD
qPCR Efficiency & Ct Reference Card
CHEAT SHEET
Nuclear Extraction Protocol
The brand ≠ The manufacturer
When you switched from supplier A to supplier B, you assumed you were buying a different product. Two different companies mean two different reagents, right?
Wrong! The same underlying antibody can be sold by more than one supplier under different brand names, meaning apparently “different” listings may actually refer to the same reagent. So when your first antibody fails and you “try another brand,” supplier B may be offering the same clone or source product as supplier A.
If so, changing brands may not amount to testing a genuinely independent antibody. You may instead be testing another lot or formulation of the same underlying reagent under a different label.
Manufacturing sources can change too
To complicate matters further, provenance can also be less stable than it looks. In some cases, a supplier’s source may change over time, or the origin of a listed product may simply not be clear from the page in front of you.
However, it’s important to keep in mind that the extent of this risk depends on the supplier’s arrangements, and unclear provenance is not, by itself, evidence of deliberate concealment.
Why provenance matters for reproducibility
Reproducibility failures rarely have a single cause. Antibody specificity, application suitability, protocol conditions, sample handling, biological variability, and the quality of reporting in the original paper all contribute.
Provenance can also be included in that list because when the line back to the producer is unclear, several of the other reproducibility factors can be affected. It is helpful to remember that:
- A lot number is only as useful as the supplier’s ability to link it to production and QC records.
- You may not be able to confirm whether a later order has the same source and specification as an earlier one.
- Validation images on the datasheet may not be clearly linked to the lot you received.
Together, these uncertainties make a project harder to diagnose because the identity of the reagent in the packaging remains unclear.
Isn’t that what a datasheet is for?
A good datasheet gives you genuinely useful information: target, clone, host, reactivity, the applications it’s been tested in, a recommended dilution range, storage and formulation, and validation images.
Datasheets can also carry provenance, lot-specific validation, clone details, an RRID, QC methods, and manufacturer information. But the problem is that many don’t present all of this clearly.
Where the datasheet can trip you up
When a datasheet lists “Western blot, IHC, flow cytometry” under applications, it’s tempting to read that as evidence the product works in each. However, it may not be.
Suppliers use application labels inconsistently: a listing might mean the product was actually tested in that application, or that it’s expected to work but hasn’t been tested. When suppliers publish negative validation (i.e., stating outright that a product should not be used for a given application), that information is useful too.
It’s also worth noting that sequence homology is more commonly used to predict likely species reactivity than to predict performance in a particular application, so a homology-based claim isn’t the same as application evidence.
The takeaway is that “application suitability” requires application-specific evidence. Validation in one application doesn’t automatically carry over to another, and it’s easy to assume it does.
An example: Application ≠ Tested
A standard denaturing Western blot typically uses SDS and heat, often with a reducing agent, so your target protein is detected in a denatured state. But an antibody that performs well under these conditions may or may not recognize the same target in a native or fixed context. Good Western blot performance does not establish that the same antibody will work in flow cytometry or another application.
Instead, the evidence you rely on needs to match your target, application, and sample type. You may be able to source this evidence independent of the supplier:
- YCharOS independently characterizes commercial antibodies using knockout-based controls and publishes its data openly.
- The Only Good Antibodies community, established by researchers at the University of Leicester, brings researchers and other stakeholders together to improve antibody-based research.
- EuroMAbNet’s antibody-validation guidance recommends checking antibody clone names or product codes, and, for polyclonal antibodies, comparing details such as the host species, immunogen, and validation images.
Resources such as these can complement supplier-provided evidence.
Consider the true cost of ambiguity
Price is one reason you may be tempted to accept thinner reagent documentation. Yes, you save money by taking a cheaper antibody with less evidence. But if it doesn’t work, you’re now buying replacements, spending more staff time, and delaying a potentially publishable result.
The truth is the best antibody purchase is the option with evidence that is sufficiently complete and relevant to your use, wherever it sits on price. In this case, a cheaper antibody with strong application-specific evidence may be a better choice than a mid-priced product. What you’re really trying to avoid is ambiguity, because ambiguity can lead to unnecessary re-buys.
Budget pressure is real, especially in academia, and choosing on price is a reasonable human response. The aim is to consider the evidence alongside price and reduce the risk of avoidable repeat purchases and delays.
Five checks before you buy
If you cannot establish what you are buying, you cannot confidently determine whether two purchases represent the same underlying reagent. So instead, you need to ask yourself what you are really buying.
To avoid ambiguity, run these five checks on a reagent before you spend your hard-earned grant money. You won’t always get every answer because some manufacturing arrangements are confidential. In that case, consider whether the available documentation, validation, and technical support are sufficient for your work.
1. Traceability: Who actually made it?
Ask the supplier who produced the antibody, which lot you’ll receive, and what QC sits behind that lot. Also ask about supply continuity if you need consistency across a long project.
Note that an ISO 9001 certificate indicates a company operates a quality-management system; it does not by itself prove the company manufactured a specific antibody, validate that antibody’s scientific performance, or guarantee it’s suitable for your application.
2. Evidence fit: Does it match your work?
Look for evidence in your application, against your target, in your sample type, not general evidence. Ask for the clone or sequence identity where available, an RRID, immunogen details, and lot-specific validation. If your target has very few available antibodies, you may need to reconsider the experimental approach or investigate a custom reagent, while accounting for the additional validation that route requires.
3. Controls: Put them in first.
Controls belong in the experimental design, not as an afterthought during troubleshooting. Decide on your positive and negative controls up front, and retain appropriately stored aliquots of a known-positive sample from an early run. If a later lot underperforms, you can use that sample to help distinguish a reagent problem from a change in the sample or protocol.
4. Lot control: Document every lot.
Record the lots you buy and compare results across them. If the same control sample performs with lots 1 and 2 but not lot 3 under matched conditions, lot 3 becomes a leading suspect. You can then investigate that possibility directly rather than assuming that the entire protocol has failed. For regulated or translational workflows, applicable quality systems may also require formal lot qualification and documentation.
5. Accountability: Who answers when it fails?
Favor a supplier who can actually resolve a technical query and point to independent characterization. Direct access to the producer may shorten that conversation when its technical team can access lot-specific QC, but a distributor with strong documentation and responsive technical support can serve you just as well.
A note on the supply chain
It’s worth picturing the path your reagent travels. At one end is the producer who made it; from there it may reach you directly, or via one or more distributors.
A longer or less transparent chain doesn’t necessarily mean the product was mishandled, as some distributors ship directly from the manufacturer or a central warehouse.
But it can increase handling complexity, make storage conditions harder to trace, and make lot-specific QC information harder to obtain. Where you can establish that a reagent is handled well and its provenance is clear, the number of intermediaries matters less than the transparency of the arrangement.
Manufacturer status alone doesn’t settle any of this either. Being the producer doesn’t mean an antibody is good, and being a distributor doesn’t make a supplier unaccountable. Validation quality, clear provenance, sensible supply agreements, lot information, and responsiveness matter more than whether the seller is also the manufacturer.
What provenance checking can’t do
One caveat worth stating plainly is that establishing provenance doesn’t establish specificity, sensitivity, suitability for your model, lot-to-lot consistency, or experimental success. It makes failures easier to trace, but it doesn’t replace validating the antibody in your own application and sample system. So make sure you don’t skip that step.
Where to put the effort
If you take one thing from this, understand that some avoidable uncertainty is introduced at the point of purchasing any reagent. Establish what you can about the reagent’s identity and source, match the evidence to your application, define your controls before starting, and record every lot. Then validate the antibody in your own system, because published validation reflects specific experimental conditions.
Unreliable reagents will not disappear, but better purchasing records and application-specific validation can make some failures easier to prevent, diagnose, and report.
You made it to the end—nice work! If you’re the kind of scientist who likes figuring things out without wasting half a day on trial and error, you’ll love our newsletter. Get 3 quick reads a week, packed with hard-won lab wisdom. Join FREE here.
