A principal component analysis (PCA) plot identifies the major components in a set of data. Principal components are groups of features in the data that help the most to distinguish one sample from another. The Proteome Discoverer application automatically calculates these principal components.

A principal component analysis displays the correlation between multivariate data in a set of observations. It transforms a set of observations of possibly correlated variables into an artificial set of independent linear combinations of the original variables known as principal components (PC1, PC2, PC3, and so on). PC1 has the most variation, and the highest principal component has the least variation.

The principal component analysis plot has three pages:

  • The Scores Plot page—Shows the correlation among the observations. For details, see Scores Plot.
  • The Loadings Plot page—Shows the relations among the variables for a given pair of principal components. For details, see Loadings Plot.
  • The Variances Plot page—Shows the proportion and the cumulative proportion of the variance contributed by each principal component. For details, see Variances Plot.

The result report must contain multiple quantification values for you to display this plot or run a new differential analysis.

Procedure

  1. Open the result file of interest.
  2. Select the Proteins or Peptide Groups tab.
  3. Select View > Distribution Charts, or select Distribution Charts in the toolbar.
  4. Select the PCA Plots tab.
  5. (Optional for the scores plot only) In the Group By area at the far left, select one or more of the following checkboxes to create sample groups according to the study factors for quantification:
  6. Study_factors: Lists each study factor that you defined in the study.
  7. Sample: Selects the sample mixture of all samples. The Sample checkbox is based on the Sample Type default study variable on the Grouping & Quantification page. It appears whether or not you select Sample Type on the Grouping & Quantification page.
  8. (Optional for the scores plot and the loadings plot only) In the Filter By area at the far left, set one or more of the following variables to filter the data:
  9. Set the Study_factor filter to On.
  10. Set the Sample filter to On.
  11. Select the arrow to the left of the On/Off filter box for Study_factor or Sample.
  12. Select the appropriate checkboxes below Study_factor, Sample, or both.
  13. From the Data Source list, select the result category to plot the data from, either Proteins or Peptide Groups, and peptide isoforms.
  14. In the X Data list, select an appropriate principal component.
  15. In the Y Data list, select an appropriate principal component.
  16. (Optional) Select the Center and Scale checkbox when you want to mean-center the values in a covariance matrix.
  17. If you clear this checkbox, the chart uses the original values in the covariance matrix.
  18. (Optional) Select the Use Normalized Abundances checkbox when you want the chart to display normalized abundance values.
  19. This checkbox, selected by default, is only available when you normalize the data with the normalization parameters of the Reporter Ions Quantifier node or the Precursor Ions Quantifier node in the consensus workflow. If you clear this checkbox, the chart uses abundance values that are not normalized.
  20. Select the tab of the type of plot that you want to view: Scores Plot, Loadings Plot, or Variances Plot.