Automated quality control method and system for genetic analysis

• US 7,398,171 B2
• Filed: 06/30/2006
• Issued: 07/08/2008
• Est. Priority Date: 06/30/2005
• Status: Active Grant

• Current Assignee: Applied Biosystems LLC (Thermo Fisher Scientific Incorporated)
• Original Assignee: Applera Corporation (Thermo Fisher Scientific Incorporated)
• Inventors: Subramaniam, Yerramalli, Woo, David C.
• Primary Examiner(s): Cosimano; Edward R
• Application Number:
  US11/428,048
• Publication Number:
  US 20070010951A1
• Time in Patent Office: 739 Days
• Field of Search: 382/100, 382/110, 382/128, 382/129, 382/130, 382/133, 702/19, 702/20, 702/81, 702/82, 702/84, 702/189
• US Class Current: 702/82
• CPC Class Codes: G16B 25/00 ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
• CPC Class Codes: G16B 25/10 Gene or protein expression profiling; Expression-ratio estimation or normalisation
• CPC Class Codes: G16B 30/00 ICT specially adapted for sequence analysis involving nucleotides or amino acids
• CPC Class Codes: G16B 45/00 ICT specially adapted for bioinformatics-related data visualisation, e.g. displaying of maps or networks
• CPC Class Codes: G16B 50/00 ICT programming tools or database systems specially adapted for bioinformatics
• CPC Class Codes: G16B 50/30 Data warehousing; Computing architectures
• CPC Class Codes: G16H 10/20 for electronic clinical trials or questionnaires
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records

Abstract

Aspects of the present invention describe a method and apparatus for automating quality control for gene expression data. A computer based device receives gene expression data associated with a spectral species and genetic sample in each well of a plate. Gene expression data may be received from a sequence detection instrument performing one or more gene expression related operations for each of the wells of the plate. The computer based device identifies gene expression data determined to have anomalous characteristics according to a set of one or more quality control metrics and may conditionally flag one or more wells of the plate affected by the anomalous characteristics. Filters can then be selectively applied to temporarily or permanently remove the flagged data from subsequent gene expression studies.

Claims

1. A computer implemented method of automating quality control for genetic analysis data, comprising:

receiving genetic analysis data associated with at least one spectral species and genetic sample in each well of a plate;

performing one or more genetic analysis related operations on the received genetic analysis data derived for each of the wells of the plate;

identifying at least one of the results from performing the one or more genetic analysis related operations determined to have anomalous characteristics according to a set of one or more quality control metrics; and

flagging one or more wells of the plate affected by the anomalous characteristics according to the set of one or more quality control metrics and making the resulting flagging of the one or more wells available for further processing.

2. The method of claim 1 further comprising:

flagging one or more studies using the plate determined to have at least one or more wells flagged for the anomalous characteristics according to quality control metrics, wherein each of the one or more studies uses a set of the one or more wells of the plate.

3. The method of claim 1 wherein the genetic analysis may be selected from a set including:

gene expression and genotyping.

4. The method of claim 1 further comprising:

determining whether to remove selected data associated with the flagged wells based upon the quality control metrics;

excluding selected flagged data from genetic analysis related operations and indicate flagged wells.

5. The method of claim 3 wherein indicating the flagged wells further includes:

displaying a first symbol on a visual graphical user interface representation of the plate indicating a well in the plate has not been flagged as anomalous according to one or more quality control metrics; and

displaying a second symbol on a visual graphical user interface representation of the plate indicating a well in the plate has been flagged as anomalous according to one or more quality control metrics.

6. The method of claim 1 further comprising:

selecting a set of one or more quality control metrics for automating the quality control of the genetic analysis according to a type of assay and experiment being performed.

7. The method of claim 6 wherein the assay and experiment being performed is selected from a set including:

allelic discrimination, absolute quantitation and relative quantitation genotyping.

8. The method of claim 1, wherein the set of one or more quality control metrics is selected from a set of quality control metrics including:

a sensor saturation analysis, a bad passive reference analysis, missing data analysis, an empty well analysis, a large composite signal error analysis, a low laser power analysis, a non-amplified well analysis, a non-amplified plate analysis, a noise spike analysis, a high-relative noise analysis, a distance between clusters analysis, Ct failure analysis, a number of clusters analysis, exponential region failure analysis, a Hardy Weinberg analysis, a proportion of outliers analysis, a small sample number in cluster, baselining failure analysis, and thresholding failure analysis.

9. The method of claim 8, wherein the sensor saturation analysis, further comprises:

determining if a measured intensity level for one or more wells in the plate exceeds a maximum signal response level of a sensor; and

indicating that the measured intensity from the one or more wells in the plate has exceeded the maximum signal response level of the sensor in response to the determination and may not reflect an accurate measurement.

10. The method of claim 8 wherein the bad passive reference analysis further comprises:

identifying a start intensity and a subsequent intensity for a passive reference in one or more wells in the plate after a number of thermal cycles;

determining if a differential between the start intensity and the subsequent intensity for the passive reference in the one or more wells exceeds a predetermined threshold; and

indicating that the one or more wells in the plate may have failed as the intensity of the passive reference should not vary beyond the predetermined threshold after the number of thermal cycles.

11. The method of claim 8 wherein the missing data analysis further comprises:

identifying a missing data point for one or more probe dyes and passive reference in one or more wells in the plate during a thermal cycle; and

indicating that the one or more wells in the plate have missing values for the given thermal cycle may be due to either a failure in system behavior or a mismatch in a configuration.

12. The method of claim 8 wherein the large composite signal error analysis further comprises:

generating an expected probe dye component contribution derived from a measured composite signal response from probe dyes and a passive reference in each well of the plate;

determining an expected composite signal response using each of the expected probe dye component contribution in conjunction with associated dye component calibrations; and

indicating that the probe dye component contribution in the one or more wells in the plate may be incorrectly setup when difference between the measured composite signal response and the expected composite signal response exceeds a predetermined error analysis value.

13. The method of claim 8 wherein the low laser power analysis further comprises:

reviewing power level data for powering a light source on an instrument during a run of an experiment;

determining when the power level has dropped below a predetermined acceptable level of driving the light source on the instrument; and

indicating that the experimental results gathered when driving the light source on the instrument may be incorrect as the power level for operating the light source may have been insufficient.

14. The method of claim 8 wherein the noise-spike analysis further comprises:

receiving an amplification curve of one or more probe dyes derived from a well in the plate;

computing standard deviation and mean noise values for a noise portion of a signal in the amplification curve;

measuring a number of standard deviations that a selected noise point differs from the mean noise value;

determining when the selected noise point on the amplification curve exceeds a predetermined standard deviation threshold compared with the mean noise value for the amplification curve; and

indicating that one or more of the selected noise points may reflect a noise spike in the amplification curve and may not reflect a typical noise point on the amplification curve.

15. The method of claim 8 wherein the high relative noise analysis further comprises:

receiving an amplification curve of a first probe dye from each well of a plate derived through multicomponent analysis;

computing a standard deviation and mean value for noise compared with the first probe dye in each well of the plate;

computing a standard deviation and mean for the noise in the plate based upon the mean values and standard deviations of the noise for the first probe dye in each well of the plate;

determining when a standard deviation of noise for the first probe dye in each well of the plate exceeds the mean value of the plate by a threshold number of standard deviations for the plate; and

indicating that the one or more wells in the plate may not accurately reflect the amplification characteristics of the probe dye due to a relative amount of noise in the one or more wells compared with the plate.

16. The method of claim 8 wherein the distance between clusters analysis further comprises:

identifying a location of a sample and a no template control (NTC) cluster on an allelic discrimination plot;

determining when the location of the sample is within a threshold distance from the location of the NTC cluster on the allelic discrimination plot; and

indicating that one or more values associated with the sample near the NTC cluster may not amplify well and therefore may be unreliable.

17. The method of claim 8 wherein the Ct failure analysis further comprises:

performing an automatic determination of Ct based upon an amplification plot of a probe dye;

determination if at least one or more regions selected from a set including an exponential region and a baseline region of the amplification plot cannot be reliably determined; and

indicating the automatic determination of Ct should not be relied upon due to the failure to accurately identify one or more regions of the amplification plot.

18. A computer program product for automating quality control for genetic analysis data, tangibly stored on a computer-readable medium, comprising instructions operable to cause a programmable processor to:

receive genetic analysis data associated with at least one spectral species and genetic sample in each well of a plate;

perform one or more genetic analysis related operations on the received genetic analysis data derived for each of the wells of the plate;

identify at least one of the results from performing the one or more genetic analysis related operations determined to have anomalous characteristics according to a set of one or more quality control metrics; and

flag one or more wells of the plate affected by the anomalous characteristics according to the set of one or more quality control metrics and making the resulting flagging of the one or more wells available for further processing.

19. The computer program product of claim 18 further comprising instructions that:

flag one or more studies using the plate determined to have at least one or more wells flagged for the anomalous characteristics according to quality control metrics, wherein each of the one or more studies uses a set of the one or more wells of the plate.

20. The computer program product of claim 18 further comprising instructions that:

determine whether to remove selected data associated with the flagged wells based upon the quality control metrics;

exclude selected flagged data from genetic analysis related operations and indicate flagged wells.

21. The computer program product of claim 18 wherein the genetic analysis may be selected from a set including:

gene expression and genotyping.

22. An apparatus for automating quality control for genetic analysis data, comprising:

means for receiving genetic analysis data associated with at least one spectral species and genetic sample in each well of a plate;

means for performing one or more genetic analysis related operations on the received genetic analysis data derived for each of the wells of the plate;

means for identifying at least one of the results from performing the one or more genetic analysis related operations determined to have anomalous characteristics according to a set of one or more quality control metrics; and

means for flagging one or more wells of the plate affected by the anomalous characteristics according to the set of one or more quality control metrics and making the resulting flagging of the one or more wells available for further processing.