Simoa™ Troubleshooting

Troubleshooting

Investigate the root cause of missing result outputs from data analysis. For troubleshooting outside the scope of data analysis, see the Simoa HD-1 Analyzer Instrument Troubleshooting Guide (USER-104-19).

01. Troubleshooting H3

01.01 — Troubleshooting Flowchart

Use this flowchart to identify and navigate to a troubleshooting topic. If you have a specific error message, see Sample Error Messages on page 10. If troubleshooting results in a rare case outcome, see the Rare Case Troubleshooting Flowchart on page 2.

close

01.02 — Rare Case Troubleshooting Flowchart

Use the Rare Case Troubleshooting Flowchart to determine when you should contact Quanterix™ Customer Support.

close

01.03 — Not a Number (NaN) and Where to Find It

Not A Number (NaN) is an undefined output that is not available. The Simoa HD-1 Analyzer uses NaN to represent missing outputs due to (1) a value needed to calculate the output is missing, (2) the conditions that the output requires are not met, or (3) the calculation result is undefined. There are many instances where NaN replaces a result output, as listed below (the name in italics is the header name for each result output in the Run History Table).

  • Concentration – Replicates Conc.
  • AEB – Replicates AEB
  • Isingle – Isingle
  • fon – Fraction On
  • Ibead – Ibead

For more information on why a particular result output is NaN, see 01.04 below, where each result output is described in more detail.

To Determine if a Result Output is NaN

  1. Go to the History and Reports tab in the Simoa HD-1 Analyzer software.
  2. Touch Run History.
  3. Find the column with the desired result output, for example, Replicates Conc. for concentration.
  4. If the column is not in view, use the horizontal scroll bar to navigate through the columns.
  5. If the column cannot be found, touch Configure Columns to add the result output to the columns.

For more information on how to add or edit displayed columns, see the Simoa HD-1 Analyzer User Guide.

01.04 — What to Do When You Have NaN as a Value

If Concentration is NaN

The Replicate concentration is calculated using the calibration curve and AEB. Since the calibration curve describes the relationship between concentration and AEB, concentration can be determined given an AEB value. Without an AEB value, the concentration cannot be obtained. There are instances in which an AEB value in range will return a NaN for concentration. In this case, there can be other problems with the calibration curve (see Calibration Curve Issues in section 01.05 below) such as the calibration curve not being calculated.

If AEB is NaN

  • If fon or Ibead is NaN, AEB is always NaN unless the majority rule is undefined. (see Q 7 What is the Majority Rule? in the Q & A section)
  • If Isingle is NaN, AEB is NaN only for analog samples because the calculation of AEB requires the Isingle value (see If Isingle is NaN)

If Isingle is NaN

Isingle is NaN when the following calculation conditions are not met:

  • At least one result with fon < 0.2
  • Other plexes in the same job all have fon < 0.7
  • Total number of active beads > 150 for results with fon < 0.2
    • The number of active beads per result is the number of beads of that plex multiplied by the fon

The weighted batch Isingle value is plex-specific and assay-specific. For example, imagine a user runs a batch containing two assays because they want to compare the performance between the assays:

  • A 2plex assay with plexes IL-6 and TNFa
  • A 3plex assay with plexes IL-6, TNFa, and GM-CSF

In this batch, 5 different Isingles will be calculated:

  • 2plex IL-6
  • 2plex TNFa
  • 3plex IL-6
  • 3plex TNFa
  • 3plex GM-CSF

In a different example, if a user runs a batch containing two homebrew assays of homebrew conditions A and B of the same analyte, then a unique Isingle is calculated for condition A and condition B. It is possible for one of the homebrew conditions to have a calculated Isingle, but the other condition to have too few active beads with a missing Isingle.

If fon or Ibead is NaN

fon gives the fraction of beads that have enzyme activity. Ibead gives the average intensity of beads with enzyme activity. A failure in image analysis is most likely the cause of NaN for either of these values. To learn more about image analysis errors, see Image Analysis Issues in section 01.05 below.

If Mean, SD, or CV for AEB or Concentration is NaN

Replicate statistics require at least two replicates to calculate the mean, standard deviation (SD), and coefficient of variation (CV). To calculate values for the replicate concentration, AEB cannot be NaN or undefined. When calculating AEB statistics, all NaN values for the batch are ignored; however, out of range Mean AEBs are included.

01.05 — Determining Calibration Curve or Image Analysis Problems

Calibration Curve Issues

Calibrators are run with the instrument to produce data points. A method of curve fitting is selected and used to fit the data points and to generate the calibration curve. Using the calibration curve, concentration is determined given an AEB value. AEB and concentration values must be within a valid range of the calibration curve, as shown in the following figure.

To generate a calibration curve, all calibrator levels defined for the assay must be present and the number of points must meet the minimum requirement of the selected fitting algorithm.

Image Analysis Issues

Image analysis issues can prevent the software from calculating analyte concentrations. These problems often result in a fon of NaN or no active beads. To view the error messages related to image analysis failure, review the Event Logs (see Using the Event Log to Investigate Error Messages in section 01.07 below).

A common problem is too much fluorescence for image analysis. One of the following error messages will appear:

  • "Too much fluorescence in resorufin channels for analysis. Check the detector and RGP reagents; if there is no problem, please dilute the source and rerun."
  • "Too much fluorescence in the resorufin channels for analysis."

This error signals that the concentration is limited by saturation of system components, such as the imaging system or substrate amount in the wells. If you are running a homebrew assay with failing calibrators, check that the detector and RGP concentrations are reasonable. When running a Quanterix kit assay or a homebrew assay with no issues analyzing calibrators, this error signals a specimen has a concentration that is too high for analysis. In this case, dilute and rerun the sample.

If you receive the "Too much fluorescence" error message for a calibrator or control in a Quanterix kit assay, or if you receive any other error messages, contact Quanterix Customer Support for troubleshooting help.

Another common error message you may see when running specimens or homebrew assays is too few beads. This error message will appear as, "Too few beads loaded for analysis to proceed". Two of the most common causes are poor bead resuspension and sample matrix effects. Make sure your beads are properly resuspended prior to running any assay. If you are uncertain about how to resuspend beads, please see the Simoa User Guide for more information. If you repeatedly see too few beads messages, especially with a particular specimen, try diluting the specimen further to reduce any matrix effects. Do not hesitate to contact your field application scientist or Quanterix support if you have any questions.

If you ever see an error message besides "Too Much Fluorescence" or "Too Few Beads", please contact Quanterix Support using the information provided at the start of this guide.

01.06 — How to Find Error Messages for a Batch or Job ID

Using the Event Log to Investigate Error Messages

The Event Log is useful when investigating unusual result outputs or errors from a run. You can also use it to determine why NaN was assigned for a result output.

To View the Event Log of One Job ID

  1. Go to the History and Reports tab.
  2. Ensure that automatic replicate selection is off.
  3. Select the job of interest.
    Note: When viewing Event Log entries of a single job, select only one item at a time. To view the Event Log entries of multiple jobs, see To View the Event Log of Multiple Job IDs or Batches below.
  4. Touch Show Related Flags and Events.

An example of an Event Log output with error messages is shown on the next page. This particular example shows that the calculated concentration and the given AEB were out of the calibration range.

To View the Event Log of Multiple Job IDs or Batches

  1. From the Run History tab, note the Job ID, Batch ID, or Creation Date of the batch that you want to investigate (other fields will work as well).
  2. Touch Event Log.
  3. Touch the Add Filters button (+) to add a filter to the displayed error messages.
  4. Use the Job ID, Batch ID, or Creation Date for the filter. See the Simoa HD-1 Analyzer User Guide for more details on how to filter results.

To Export Error Messages

  1. Note Test ID, Sample Barcode, Creation Date, or Completion Date.
  2. In the History & Reports tab, touch Reports.
  3. Select Exceptions Report from Report Types.
  4. Touch the Add Filters button (+) to add a filter to the displayed error messages.
  5. Select the check boxes of the messages to export.
  6. Touch Preview Report and wait for the report to show.
  7. Touch Export and specify the destination of the file.

Sample Error Messages

The Sample Error Message table lists error messages that can appear in the Event Log.

MessageDescription
Could not calculate AEB across replicates. Check the event log for possible individual replicate error messages.Check the event log of the individual replicates to troubleshoot.
Unable to correct for optical signal interference.There was a problem with image analysis which did not allow analyte concentrations to be calculated. Contact Quanterix Customer Support for more information.
Could not calculate the required ISingle value.The Isingle value was not calculated or is undefined. Check individual replicates for failure. In a Homebrew assay, ensure you followed guidelines for Isingle calculation.
Too few beads loaded for analysis to proceed.There were not enough active beads in the batch or plex. In a Homebrew assay, check bead concentrations. Otherwise, contact Quanterix Customer Support for more information.
Too few valid wells found for analysis.Contact Quanterix Customer Support.
Too much debris found in the image to proceed with image analysis.Excessive debris prevented image analysis from completing. Contact Quanterix Customer Support.
Too much fluorescence in re-sorufin channels for analysis.Check the detector and RGP reagents. If no problem is found, dilute the source and rerun. Excessive fluorescence prevented image analysis from completing. See Image Analysis Issues in section 01.05 for more details. TIP: Diluting and rerunning the samples can decrease resorufin activity and fluorescence.
Image analysis failure due to subimage misalignment.Image analysis could not complete because the sample images could not be aligned. Contact Quanterix Customer Support.
Unable to match input bead types to image.An exception was raised because image analysis could not determine which wells were beaded and which wells were not. Contact Quanterix Customer Support.
Invalid AEB numeric value found for a calibration data point.A calibrator level value was reported as NaN. To create a calibration curve, all calibrator levels must have values. Manually excluding calibrator levels with NaN values will allow for curve creation. Check the individual replicates for failure reasons.
Calculated Concentration X is out of calibration range for the current curve.The calculated concentration is either above the highest calibrator concentration or below the lowest calibrator concentration.
Could not calculate calibration curve, because the given number of data points is insufficient.The calibration curve could not be fit into the data points because there were not enough points. The required number of data points depends on the algorithm used to fit a curve.
Sample AEB value, X, is outside of the range established by the calibration curve.A NaN value is displayed for the calculated concentration. No replicate results are calculated. Mean, SD, CV, AEB, and concentration values are not available.
A valid calibration curve fit does not exist for current plex.The calibration curve for a given plex was not found or could not be created. In a Homebrew assay, ensure your curves are set up correctly. Otherwise, check the individual failures.
Analog AEB unavailable for plex X.Analog AEB is unavailable because Isingle is NaN. See If Isingle is NaN in section 01.03 for information on troubleshooting Isingle issues.
Insufficient volume detected by probe.The software calculates how much volume is left based on the geometry of the container and compares it against how much is to be aspirated. If the volume available is less than the volume to aspirate, this error appears.
Job canceled. No cuvette available for job.The job is canceled due to missing cuvettes (or cuvettes that the software is expecting). This situation can occur if (1) The user forgets to load the first 10 cuvettes before the start, or (2) A scheduler issue prevents the system from providing the needed cuvettes.
Job canceled. No liquid level detected.To prevent physical damage, each pipettor has a maximum Z position at which the pipette stops moving. If no liquid is detected when the pipette moves down to aspirate for a job at the maximum position, the job is canceled.
Need at least two valid results for this plex in order to calculate a replicate result.Mean, SD, or CV cannot be calculated because there are fewer than two replicate results.
No cuvette available for job.See Job canceled. No cuvettes available for job described earlier in this table.
No liquid found by probe.See Job canceled. No liquid level detected described earlier in this table.
System stopped and job canceled.System halts (for any reason) and all unfinished jobs are canceled.

01.07 — Frequently Asked Questions

Questions and Answers

01 When is the Replicate AEB out of range? Replicate AEB is out of range when it is above the highest calibration level or below the lowest calibration level. The AEB concentration figure shows the highest and lowest calibration level (see Calibration Curve Issues on page 5). The concentration will still be calculated.
02 When is the Concentration out of range? Concentration is out of range when it is above the concentration of the highest calibrator or below the lowest calibrator.
03 AEB was out of range but I got a concentration. What is happening? If AEB is out of range, the software will calculate the concentration but will flag the result.
04 What are the conditions for the calibration curve to be created? For the calibration curve to be generated, all calibrator levels defined for the assay must be present and the number of points must meet the minimum requirement of the selected fitting algorithm.
05 What are the conditions for the Mean AEB to be calculated? To calculate the Mean AEB, you need at least two replicates for samples that are run in replicate.
06 If the replicate AEB is out of range, is it still included in the calculation of Mean AEB? No, the calculation of Mean, SD, and CV of AEB does not include AEB values that are out of range.
07 What is the majority rule? For replicate samples, the majority rule is a method to determine whether samples are treated as analog or digital. It counts the number of analog or digital samples in a batch and uses the method with the majority.
08 When is the analysis mode undefined? The analysis mode is undefined when the Fraction On or Digital AEB is NaN. It can also be undefined if the sample is run in replicate and all but one of the replicate has a job that is canceled.
09 Are samples with AEB out of range included in the majority rule? Yes, to calculate sample AEB, the majority rule must be applied first. Determination of whether AEB is out of range takes place after AEB is calculated, thus after the majority rule is applied.
10 Are the Mean, SD, and CV for AEB and concentration calculated for specimens that have one or more NaN values? Yes, Mean, SD, and CV are calculated from replicates that do not have NaN values.
11 I have all the calibrator values but the software did not return a curve. This could be a problem with a severe outlier that causes the curve fit to fail. In such cases, exclude the outlier and perform the curve fit again.
12 What happens when the number of analog and digital replicates is the same? When the number of analog and digital replicates is the same, the majority rule goes into a tie breaker and defaults to the analog method.
13 Majority rule could not be extracted. What does this mean? When the majority rule cannot determine if an analog or digital analysis method is used, an error message appears indicating that the majority rule could not be extracted. This is because all of the replicates have undefined analysis methods.
14 How is the majority rule evaluated when there is a manual exclusion/inclusion of replicate samples? Manual exclusion/inclusion of replicate samples forces the majority rule to be re-evaluated.
15 What is the difference between the analog and digital analysis methods? The digital analysis method uses fon to calculate the value of AEB. The analog analysis method requires fon, Ibead, and Isingle to calculate the AEB value. The digital method yields better precision when the AEB is lower than 1.2, corresponding to an fon of 0.7, whereas the analog method is better when the AEB is higher than 1.2.
16 How is a sample determined to be analog or digital? A sample's majority rule vote is digital when it's fon, the fraction of beads with active enzymes, is less than 0.7, otherwise it is analog. Based on the votes from all sample replicates, a sample is either analog or digital.
17 What determines if the sample is treated as analog or digital? The majority rule is used to determine if a sample is analog or digital.
18 Why would a batch require at least one sample with fon < 0.2? If the batch contains analog samples, an Isingle value is required to calculate AEB values for these samples. Isingle can only be calculated if samples with fon < 0.2 are present in the batch.