Blog

Quick answer A researcher measuring circulating lipopolysaccharide (LPS) in human serum with our LPS ELISA saw almost no color, even after 30 minutes with substrate. Samples had been tested neat and at 1:10, 1:100 and 1:1000, with no standard curve on the plate. A second run with the full standard curve gave an excellent curve, which confirmed the kit was working. Most samples still read at or below the blank. Circulating LPS is usually very low, often in the tens of pg/mL, so any dilution pushes it below the 7.8 pg/mL lowest standard. Run serum undiluted, always include the standard curve, and report values below the curve as “< LLOQ”. On this pag …

Quick answer A researcher running an 8-OHdG ELISA and a creatinine ELISA on human urine got 8-OHdG/creatinine ratios in the hundreds, then the thousands, of ng/mg, far above the ~30–40 ng/mg reported in the literature. Two errors in the calculations came first: a curve-fit equation that was never solved for concentration, and a poorly fitting model. Once those were fixed, the calculations were sound. The ratios still looked high because neither analyte had been diluted into the middle of its own standard curve and checked for dilution linearity in urine. In a ratio, both of those errors push the result in the same direction. On this page Th …

Quick answer A researcher measuring glucose in human saliva on the fluorometric procedure of a glucose assay kit had a textbook standard curve, yet every sample read far below the expected physiological range — unfiltered samples worst of all. The kit was not faulty: undiluted saliva is a complex matrix that suppresses the reaction (matrix interference), and the neat signal also sat at the bottom of the narrow fluorometric range. A dilution-linearity series (1:5 → 1:25) lifted the values until they plateaued at 1:20–1:25, landing in the expected 0.7–2 mg/dL band. The fix: dilute complex matrices into the middle of the curve and confirm the result i …

Quick answer A researcher ran a high-sensitivity porcine insulin sandwich ELISA on EDTA plasma and saw main-run duplicate CVs as high as 29.6% (kit spec: intra-assay CV <8%) together with a standard curve whose lowest point sat only about 0.0018 OD above the blank. Despite an R² >0.99, the curve was compressed because colour development was stopped too early (~6 minutes). Extending TMB development into the validated 10–20 minute window (~15 min, top standard reaching ≥1.5 OD), validating the curve before committing precious samples, and diluting the plasma into the middle of the curve (1:5–1:10) restored interpretable, reproducible …

Quick answer When plates from the same lot give very different OD values, the cause is almost always a comparison problem, not a manufacturing problem. The three usual culprits are a truncated standard curve (top standard omitted), non-identical run conditions (different substrate development times), and a single plate compromised by a protocol deviation. Before concluding a kit is defective, rebuild every plate on the full standard curve under identical conditions — here, a PACAP ELISA on EDTA plasma with a 7.813–500 pg/mL range. On this page The problem Snapshot of the data Why it happens Our analysis Root cause …

Quick answer A researcher measuring dopamine in mouse brain tissue got no detectable signal and, with many samples to run, asked whether the full standard curve could be skipped to save wells. It can't. The standard curve does two jobs: it confirms the reagents are working and it provides the equation that converts each optical-density reading into a concentration — without it there is simply no way to calculate a result. The likely reason nothing was detected is a matrix effect from the high-protein tissue extract, which is resolved by diluting the samples (e.g. 1:5, 1:20, 1:100) and running the complete standard curve alongside them. On this page …

Quick answer A researcher's colorimetric adenosine assay looked broken — the standard curve came out with a negative slope and the calculated values made no sense. The assay itself was working perfectly. The results came from how the data was being handled: the blank was in the wrong plate position, the two plate readings were subtracted in the wrong order, and the standards were fitted with the wrong model and the wrong units. Fixing the analysis produced clean, positive standard curves and blanks reading near zero — no re-run needed. On this page The problem The data Why it happens Our analysis Root cause What we recomm …

Quick answer A researcher validating two isoform assays — the Human ALT1 (HUFI00557) and ALT2 (HUFI00593) ELISA kits — spiked the ALT2 kit's standard onto the ALT1 kit's plate and saw a signal as strong as the ALT1 standard itself. The kits were not cross-reacting against native ALT. Both standards are tagged recombinant proteins that share a common expression tag, and the polyclonal antibodies carry a minor sub-population that recognises that tag. The cross-talk lives entirely in the tagged standards — it does not occur with native, untagged ALT in real samples. On this page The problem Snapshot of the data Why it happens Our analys …

Quick answer A researcher running a Rat Eotaxin/CCL11 (Eotaxin) sandwich ELISA saw sample readings sit above the top standard and wells turn colour the instant TMB was added — even after 1:2 and 1:5 dilutions. This was not a faulty kit. The fetal bovine serum (FBS) in the conditioned culture medium created a high background: the medium-only signal alone reached more than 60% of the top standard, so no practical dilution could bring readings into a reliable range. The fix is to quantify the matrix with a medium-only blank, then either collect serum-free conditioned medium or switch to cell lysates in an HRP-friendly buffer. On this page The probl …

Quick answer A researcher's samples read negative at the standard 1-hour incubation on the MycoGenie Rapid Mycoplasma Detection Kit (MORV0011), but many turned positive after 2 hours. This is not a false positive. The cultures were sub-confluent (~40%), so the low mycoplasma biomass kept a genuine, low-level infection below the colour-change threshold at 60 minutes. Anchoring the run with the kit's positive and negative controls and extending incubation modestly (~90 minutes) confirmed the true positives. On this page The problem Snapshot of the data Why it happens Our analysis What we recommended FAQ …

QUICK ANSWER A researcher ran a panel of EMT-marker ELISA kits (E-cadherin, claudin-1, N-cadherin, vimentin, ZEB1, TWIST1) on breast cell-line lysates and found E-cadherin reading "positive" in a mesenchymal line where it is barely expected — an apparent false positive. The kits were fine. The mismatch came from comparing raw lysate concentrations across cell lines without normalizing to total protein, plus a detergent-heavy lysis buffer the assays are not validated for. Normalize each result to total protein (analyte per mg protein), use a compatible buffer, and the expected epithelial-vs-mesenchymal pattern returns. On this page The problem …

QUICK ANSWER On a Human I-FABP/FABP2 ColorStep ELISA, the top of the standard curve looked perfect while the three lowest standards showed duplicate %CVs of 37–450% — because their optical density (OD) sat right at the blank. Near the bottom of the working range, a few milli-OD of well-to-well noise becomes a huge relative error, and dual-wavelength (OD450–620) subtraction pushed near-zero wells negative. The fix: keep the curve inside the validated 0.156–10 ng/mL range, retain single-wavelength OD450, and bring low samples up into the reliable mid-curve (run neat or concentrate). QC confirmed no kit defect. On this page The pro …

Quick answer: A researcher subtracted the blank from a competitive ELISA (a Serotonin kit) and every result turned negative. In a competitive assay the signal is inversely proportional to concentration, so the blank (zero standard) gives the highest optical density — subtracting it drives all readings below zero. Do not subtract the blank in a competitive ELISA. Blank subtraction belongs to sandwich ELISAs (like the researcher’s Zonulin kit), where signal rises with concentration and the blank is the lowest value. On this page The problem The data Why it happens Our analysis Root cause What we recommended Ke …

Quick answer On the Dog Beta-endorphin (BEP) ELISA Kit, a researcher ran the same canine serum aliquots on two plates from the same lot and saw roughly double the concentration on one plate versus the other. The standard curves on every run were valid and the lot’s QC precision was strong (inter-assay CV ≈ 6.8%), so the kit chemistry was working. The discrepancy traced to sample-level handling and an undiluted-serum matrix effect on a single run — not a kit fault. Diluting serum into the mid-range of the curve and tightening pipetting resolved it. On this page Click any section below to jump straight to it ↓ The pr …

Quick answer On the Periaxin (PRX) ELISA, a researcher found every serum sample reading above the top standard — optical density (OD) around 2.8–3.0 at 450 nm, well beyond the highest calibrator. The kit was not faulty. The samples simply contained far more analyte than a 1:6 dilution could bring into the assay’s 0.156–10 ng/mL range. The fix is to confirm the standard curve, then dilute samples enough (here, closer to 1:100–1:1000) to land them inside the linear region. On this page Click any section below to jump straight to it ↓ The problem › Snapshot of the data › Why it ha …

Quick answer On the Fatty Acid Oxidation (FAO) Assay Kit (BR00001), a researcher found that a subset of skeletal-muscle samples returned negative values once the control well was subtracted from the reaction well, while other samples on the same plate read clearly positive. The negatives were not a kit failure: they mark samples whose FAO activity sits at or below the assay's detection floor, where ordinary well-to-well optical noise tips the background-subtracted value just below zero. The fix is interpretation, not a repeat run — report these as ND (not detectable), confirm blood was washed from the tissue, and let the valid positive samples validate the plate. …

Quick answer A researcher preparing a Human IL-18 ELISA wasn't sure whether the protocol's “300-fold” HRP-Streptavidin step meant adding 40 µL of concentrate to 12 mL of diluent, or to a total of 12 mL. A fold dilution always describes the final total volume, not the diluent alone. For a 300-fold dilution you combine 1 part concentrate (40 µL) with 299 parts diluent (11,960 µL) to reach 12 mL total — so 40 µL HRP-Streptavidin + 11,960 µL Assay Diluent = 12 mL. The worked example's rounded “12 mL diluent” wording was the source of the confus …

Quick answer A competitive ELISA is running correctly when the blank (zero standard) gives the highest OD, the standard curve drops smoothly with rising concentration, and your samples land in the middle of that curve. If your interpolated values already sit comfortably inside the kit's range — for the Bovine PGE2 (PGE2) ELISA that is 15.6–1000 pg/mL — you generally do not need to dilute. Fit the standards with a 4-parameter logistic (4PL) regression and keep duplicate CVs at or below 20%. On this page The problem The data Why it happens Our analysis Root cause What we recommended Key takeaway FAQ The problem A researcher reached out before starting a new …

QUICK ANSWER If your ELISA returns sample concentrations that are orders of magnitude too high, the assay itself is usually fine — the error is almost always in data analysis. The most common culprit is entering the wrong standard concentrations when building the standard curve. In a recent porcine MAP (PigMAP) sandwich ELISA case, a researcher assigned a 5000–78 ng/mL range to standards that should have been 10–0.156 ng/mL. The optical density (OD) readings were valid; only the concentrations tied to them were wrong. Re-assigning the correct values to the same ODs brought results back into a sensible range. On this page The problem Snapshot of t …

Quick answer On the GDF-6 (Growth Differentiation Factor 6) ELISA Kit, a researcher saw an elevated blank (OD ≈ 0.41) with the two lowest standards reading at or below it, and no signal for a recombinant GDF-6 sample. The cause was a background/handling issue — most likely an over-concentrated HRP conjugate, incomplete washing, or stale wash buffer — not a fault in the antigen or kit. Fresh reagents, thorough washing, and a rebuilt low-end curve resolve it. On this page Click any section below to jump straight to it ↓ The problem › Snapshot of the data › Why it happens › Our a …