labmatetools / 5.5 ELISA
ALL TOOLS

ELISA Standard Curve Analyzer

Open or paste the plate reader export, tap the 8×12 layout, fit a weighted 4PL or 5PL curve, and read concentrations multiplied by each sample's own dilution. High replicate CVs, poor recovery and samples outside the quantification range are flagged — with how much to re-dilute. Nothing leaves your browser; no account.

1 · READINGS SECTION 05 — CELLS & PROTEIN · 5.5

EXAMPLE FILE

File not read right? Report the format and we'll fix it.

PASTE READINGS From Excel or the reader software: 8 rows × 12 columns, with or without the A–H / 1–12 labels

A two-column list (well and value, A1 or A01) works too. Decimal commas are read; overflow marks such as OVRFLW, Over or *** count as saturated wells.

REPORT A FORMAT PROBLEM File not recognised, wrong read picked, values in the wrong wells? Tell us and we'll fix the reader.
WHAT WENT WRONG
EXACTLY WHAT WILL BE SENT

        
        

2 · LAYOUT TAP WELLS: STANDARDS · SAMPLES · CONTROLS · BLANKS

TAP TO SET
REPLICATES
DIRECTION
COMMON LAYOUTS
STANDARDS
IDWELLSCONC.
SAMPLES
IDNAMEWELLSDILUTION 1 :

3 · STANDARD CURVE

MODEL
WEIGHTING
ASSAY
QUANTIFICATION RANGE
STANDARDS, READ BACK OFF THE CURVE
IDNOMINALWELLS (OD)MEAN ODCV %BACK-CALC.RECOVERYRANGE

4 · RESULTS CONCENTRATION = FROM CURVE × DILUTION

Open or paste your readings, or try the example first.
IDNAMEMEAN ODCV %IN WELLDIL.CONCENTRATION STATUSNEXT STEP

Tap a row for the working. Grey numbers are outside the quantification range — shown for reference, not for reporting.

FOR THE REPORT analyte, kit lot, operator — printed on the report and kept in the record
METHODS PARAGRAPH — check it against what you did

ABOUT THIS TOOL

How to use

  1. Export the plate from the reader software — the 8×12 plate view or a list of wells — and open it here, or paste the block from Excel. When there are several reads (450 and 570/630 nm), pick the signal and the reference to subtract.
  2. Set the layout: pick a common layout, or choose STANDARD, SAMPLE, CONTROL or BLANK, the number of replicates and the direction, and tap the first well of each. Tap EXCLUDE then a well to leave a bad well out.
  3. Type the standard concentrations (or the top and the fold), the dilution of each sample and the expected value of any control.
  4. Check the curve: the fit, the back-calculated standards and the quantification range. AUTO picks 4PL or 5PL and says why.
  5. Read the results: each flag comes with the next step. Print the report (or save it as PDF), take the CSV, and SAVE RECORD to reopen the same analysis later.

The formula

y = D + (A − D) ÷ [1 + (x ÷ C)B]G
AOD at zero concentration
DOD at saturation
Cinflection point (EC50 for 4PL)
B, Gslope; asymmetry (G = 1 is 4PL)
xC · {[(A − D) ÷ (y − D)]1/G − 1}1/B × dilution

Example. 4PL with A = 0.05, B = 1.2, C = 150 pg/mL, D = 3.0. A sample diluted 1:10 reads OD 1.20: (A − D) ÷ (y − D) = −2.95 ÷ −1.80 = 1.639; 1.639 − 1 = 0.639; 0.6391/1.2 = 0.688; × 150 = 103.3 pg/mL in the well, × 10 = 1033 pg/mL.

Good to know

  1. Free, no account, not tied to any kit or reader brand. Files are read in your browser and never uploaded; the last analysis stays on this device until you press CLEAR.
  2. Concentrations come from the mean OD of the replicates; each well's own value is in the working and the CSV.
  3. The tool never drops a well by itself and never reports an extrapolated value as a result.

使用步骤

  1. 从读板软件导出整板数据(8×12 板式或孔号列表),在这里打开,或者从 Excel 直接粘贴。如果有多个波长(450 和 570/630 nm),选好信号读数和要减去的参比读数。
  2. 设定板布局:选一个常用布局,或者选“标准品 / 样本 / 质控 / 空白”、复孔数和方向,然后点每组的第一个孔。选“排除”再点某个孔,可以把坏孔剔除。
  3. 填入标准品浓度(或者最高浓度和倍比)、每个样本的稀释倍数,以及质控的预期浓度。
  4. 检查标准曲线:拟合参数、标准品回算和定量范围。“自动”会在 4PL 和 5PL 之间选择并说明理由。
  5. 查看结果:每个标记都附有下一步该怎么做。可以打印报告(或另存为 PDF)、导出 CSV,并用保存记录在以后重新打开同一份分析。

公式

y = D + (A − D) ÷ [1 + (x ÷ C)B]G
A浓度为零时的 OD
D饱和时的 OD
C拐点(4PL 时即 EC50)
B, G斜率;不对称因子(G = 1 即 4PL)
xC · {[(A − D) ÷ (y − D)]1/G − 1}1/B × 稀释倍数

例子。4PL 参数 A = 0.05,B = 1.2,C = 150 pg/mL,D = 3.0。某样本 1:10 稀释后 OD 为 1.20:(A − D) ÷ (y − D) = −2.95 ÷ −1.80 = 1.639;1.639 − 1 = 0.639;0.6391/1.2 = 0.688;× 150 = 孔内 103.3 pg/mL,× 10 = 1033 pg/mL。

须知

  1. 免费、免登录,不绑定任何试剂盒或酶标仪品牌。文件在浏览器里读取,从不上传;最近一次分析保存在本机,按“清空”即删除。
  2. 浓度由复孔的平均 OD 反算;每个孔各自的值在计算过程和 CSV 里。
  3. 工具不会自动剔除任何孔,也不会把外推值当作结果报告。
REFERENCE · ELISA STANDARD CURVES

Exporting from your plate reader

Export the plate (all 96 values), not the protocol or experiment file. Menu names change between software versions; these are the usual routes. The layouts were written from the vendors' documented export formats.

Reader softwareExportWhat the tool reads
BioTek / Agilent Gen5 (Synergy, Epoch, ELx808, Cytation)File → Export, or Report/Export Builder → plate view, to Excel or textThe 1–12 / A–H matrix for each read (“450”, “Read 1:630”), also two reads on alternate lines; OVRFLW as saturated
Molecular Devices SoftMax Pro (SpectraMax, FilterMax)File → Export → plate format, .txtEach “Plate:” block, wavelengths side by side (e.g. 450 and 570)
Tecan i-control, Magellan, SparkControl (Infinite, Sunrise, Spark)Excel output, plate viewThe block with the <> corner, one per wavelength; “OVER” as saturated
BMG LABTECH MARS (CLARIOstar, SPECTROstar, FLUOstar)Export → Excel, plate or list view“Raw Data (450)” matrices, or the Well / Content / value list
Thermo Scientific SkanIt (Multiskan, Varioskan)Export results to ExcelThe plate matrix on each measurement sheet
PerkinElmer / Revvity EnVision, VICTORPlate results as .csvWell (A01) and Result / signal columns
Bio-Rad iMark (Microplate Manager)Export raw data, text or ExcelThe 8×12 matrix
Any readerCopy the 8×12 values from the screen or Excel8 rows of 12 values, labels optional; or a well + value list

Reader not on the list, or the plate isn't read right? Use REPORT A FORMAT PROBLEM in step 1 — it can send the file's layout with every digit masked — or the contact form. Old binary .xls files can't be read — save them as .xlsx or .csv.

Checks the tool runs

FlagWhenWhat to do
> ULOQConcentration in the well above the highest standard that recovers within limitsRe-run diluted by the fold shown, which brings it to the middle of the range
ABOVE CURVE / SATURATEDOD at or above the top plateau, or a reader overflow markThe value can't be estimated; dilute at least by the fold shown (the width of the range)
< LLOQ, BELOW CURVEBelow the lowest standard that recovers within limitsRe-run less diluted if it was diluted; otherwise report < LLOQ × dilution
< LODBelow the zero standard (or blank) mean + 3 SD (− 3 SD for competitive)Indistinguishable from zero
HIGH CVReplicate OD CV above your limit (15% by default; many kits ask ≤ 20%)Look at the wells; exclude a clearly wrong one or repeat
RECOVERYControl outside 80–120% of its expected value; standards outside 80–120% (75–125% at the ends)Check the standard dilutions, the control and the curve fit
5PL NOT TRUSTED5PL didn't converge, sits on a bound, has an asymmetry with SE larger than itself, or different starts disagree4PL is used instead; add standard levels if you need 5PL
参考 · ELISA 标准曲线

从酶标仪软件导出

导出整板读数(全部 96 个值),而不是方案文件或实验文件。不同软件版本的菜单名可能不同,下面是常见路径。这些格式是根据厂商公开的导出格式编写的。

读板软件导出工具读取的内容
BioTek / Agilent Gen5(Synergy、Epoch、ELx808、Cytation)File → Export,或 Report/Export Builder → 板视图,导出 Excel 或文本每个读数的 1–12 / A–H 矩阵(“450”“Read 1:630”),两个读数隔行排列也可以;OVRFLW 视为饱和
Molecular Devices SoftMax Pro(SpectraMax、FilterMax)File → Export → plate 格式 .txt每个“Plate:”块,多个波长并排(如 450 和 570)
Tecan i-control、Magellan、SparkControl(Infinite、Sunrise、Spark)Excel 输出,板视图左上角为 <> 的矩阵,每个波长一个;“OVER”视为饱和
BMG LABTECH MARS(CLARIOstar、SPECTROstar、FLUOstar)Export → Excel,板视图或列表“Raw Data (450)”矩阵,或 Well / Content / 读数列表
Thermo Scientific SkanIt(Multiskan、Varioskan)结果导出到 Excel每个测量工作表中的板矩阵
PerkinElmer / Revvity EnVision、VICTOR板结果导出 .csvWell(A01)和 Result / 信号列
Bio-Rad iMark(Microplate Manager)导出原始数据,文本或 Excel8×12 矩阵
国产及其他酶标仪从屏幕或 Excel 复制 8×12 读数8 行 12 列数值,标签可有可无;或“孔号 + 读数”列表

你的酶标仪不在列表里,或者板数据读得不对?请用第 1 步里的报告格式问题——它可以附上数字已遮盖的文件版式——或者用联系表单。旧版二进制 .xls 读不了,请另存为 .xlsx 或 .csv。

工具做的检查

标记条件怎么办
> ULOQ孔内浓度高于回收率合格的最高标准品按提示的倍数再稀释后重测,使其落在定量范围中段
高于曲线 / 饱和OD 达到或超过曲线上平台,或酶标仪给出溢出标记无法估算浓度;至少按提示倍数(即定量范围的跨度)稀释
< LLOQ、低于曲线低于回收率合格的最低标准品若样本已稀释,降低稀释倍数重测;否则报告 < LLOQ × 稀释倍数
< LOD低于零浓度标准品(或空白)均值 + 3 SD(竞争法为 − 3 SD)与零无法区分
CV 偏高复孔 OD 的 CV 超过设定上限(默认 15%,很多试剂盒要求 ≤ 20%)检查各孔;剔除明显异常的孔或重测
回收率质控在预期值 80–120% 之外;标准品在 80–120% 之外(两端 75–125%)检查标准品稀释、质控品和曲线拟合
5PL 不可信5PL 未收敛、参数卡在边界、不对称因子的标准误大于其本身,或不同初值结果不一致改用 4PL;如确需 5PL,请增加标准品浓度点

LAST UPDATED · HOW VALUES ARE CHECKED · ABOUT LABMATE.TOOLS

QUESTIONS · ELISA STANDARD CURVE ANALYZER

Frequently asked

Should I fit my ELISA standard curve with 4PL or 5PL?

Start with 4PL. 5PL adds an asymmetry parameter and helps when the curve bends differently at the top and bottom, but it needs more standard levels and can fit noise. The tool fits both, uses 5PL only when an F-test says it fits significantly better and its parameters are well determined, and tells you why.

Why weight the fit by 1/Y²?

ELISA replicates scatter more at high OD than at low OD. An unweighted fit lets the top standards dominate and the low end, where most samples sit, fits worst. Weighting each point by 1/Y² (or 1/Y) evens that out. The weights here come from the mean OD of each standard level (at least 1% of the OD span), so the result can be reproduced in R or Prism.

How are LLOQ and ULOQ decided?

Each standard is read back off the fitted curve. The quantification range is the widest run of neighbouring standards that recover 80–120% of their nominal concentration with a CV of at most 20%, extended by one level at either end if that level is within 75–125% and 25% — the acceptance limits for ligand-binding assays in FDA and ICH M10 guidance. ALL STANDARDS uses the lowest and highest standard instead.

What do I do with a sample above the top standard?

Don't report the extrapolated number. The tool flags it > ULOQ and tells you how much more to dilute it so it lands in the middle of the range, for example “dilute 5× more, re-run at 1:50”. A saturated well can't be estimated at all, so the advice is a minimum.

Should I subtract the zero standard?

Either fit the zero standard as a point on the curve (the default) or subtract its mean OD from every well; both are common, so follow your kit insert. If the plate has substrate-only blank wells, subtract those.

Does it work for competitive ELISA?

Yes. The curve direction is detected from the standards (signal falling with concentration), %B/B0 is shown against the zero standard, and the limit of detection is taken below the zero-standard signal.

Is my data uploaded?

No. Files are read and fitted in your browser. Nothing is sent anywhere and there is no account. The last analysis stays on this device until you press CLEAR.

常见问题 · ELISA 标准曲线分析

常见问题

标准曲线用 4PL 还是 5PL?

先用 4PL。5PL 多了一个不对称参数,曲线上下两端弯曲程度不同时更贴合,但需要更多的标准品浓度点,也更容易拟合到噪声。工具会同时拟合两种模型,只有当 F 检验显示 5PL 明显更好、且参数估计稳定时才采用 5PL,并说明理由。

为什么用 1/Y² 加权?

ELISA 复孔在高 OD 处的离散度比低 OD 处大。不加权时高浓度标准品主导拟合,而大多数样本所在的低浓度端反而拟合得最差。按 1/Y²(或 1/Y)给每个点加权可以平衡这一点。这里的权重取自每个标准品浓度点的平均 OD(下限为 OD 跨度的 1%),因此结果可以在 R 或 Prism 里复现。

LLOQ 和 ULOQ 怎么确定?

把每个标准品代回拟合曲线反算浓度。定量范围是回收率在标示浓度 80–120%、CV 不超过 20% 的最长一段连续标准品;如果两端相邻的浓度点在 75–125%、CV 25% 以内,再各向外扩展一个点——这是 FDA 和 ICH M10 对配体结合分析的接受标准。选择“全部标准品”则直接用最低和最高标准品。

样本超过最高标准品怎么办?

不要报告外推值。工具会标记为 > ULOQ,并告诉你需要再稀释多少倍才能落在定量范围中段,例如“再稀释 5 倍,按 1:50 重测”。饱和孔完全无法估算浓度,所以给出的是最低稀释倍数。

零浓度标准品要不要扣除?

可以把零浓度标准品作为曲线上的一个点参与拟合(默认),也可以从每个孔中减去它的平均 OD;两种做法都很常见,请按试剂盒说明书。如果板上有只加底物的空白孔,则扣除空白孔。

支持竞争法 ELISA 吗?

支持。曲线方向由标准品自动判断(信号随浓度下降),并以零浓度标准品为基准显示 %B/B0,检测限取在零浓度标准品信号之下。

数据会上传吗?

不会。文件在浏览器中读取和拟合,不会发送到任何地方,也不需要账号。最近一次分析保存在本机,按“清空”即删除。