PGS Documentation

Page tree

Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Principal component analysis (PCA) can be invoked on performed to visualize clusters in the methylation data to reveal clustering of the samples, but also as a quality control procedure (detection of outliers could point to possible low quality or mislabeled samples). To obtain the PCA plot, switch to the Scatter Plot tab, push Recompute ( Image Removed ) and from the Color by drop down list select HPSC. Use the Rotate Mode (Image Removed )to explore the plot from different angles, as seen in Figure 1. Each as outliers within a group could suggest poor data quality, batch effects, mislabeled samples, or uninformative groupings.

  • Select Plot PCA Scatter Plot from the QA/QC section of the Illumina BeadArray Methylation workflow to bring up a Scatter Plot tab
  • Select 3. shRNA for Color by and 2. State for Size by 
  • Select (Image Addedto enable Rotate Mode
  • Left click and drag to rotate the plot and view different angles (Figure 1)

Each dot of the plot is a single sample and represents the average methylation status across all CpG loci. The result is shown in the demonstrating We can see clear separation of naive and primed HPSC hPSCs from the cells naive hPSCs transduced with short hairpin (sh) RNA shRNA lentiviruses (shNANOG and shPOU5F1). One shNANOG sample might be an outlier, but we will not exclude it for this tutorial. 

 

Numbered figure captions
SubtitleTextPrincipal components analysis (PCA) showing methylation profiles of the study samples. Each sample is represented by a dot, the axes are first three PCs, the number in parenthesis indicate the fraction of variance explained by each PC. The number at the top is the variance explained by the first three PCs. The samples are colored by levels of

...

3. shRNA and colored by 2. State
AnchorName

...

Annotated PCA plot

...

Image Added

 

Next, distribution of M-values across the samples can also be inspected by a box-and-whiskers plot: QA/QC >

  • Select Plot Sample Box

...

  • and Whiskers Chart from the QA/QC section of the Illumina BeadArray Methylation workflow to bring up a Scatter Plot tab

Each box-and-whisker is a sample and the y-axis shows M-valuesvalue ranges. Samples in this data set seem reasonably uniform and no outliers can be detected (Figure 2).

 

Numbered figure captions
SubtitleTextBox and whiskers plot showing distribution of M-values (y-axis) across the study samples (x-axis). Samples are colored by a categorical attribute (HPSC). The middle line is the median, box represents the upper and the lower quartile, while the whiskers correspond to the 90th and 10th percentile of the data
AnchorNamebox whiskers plot

...

Image Added

 An alternative way to take a look at the distribution of M-values is a histogram (QA/QC > Plot Sample Histogram).

  • Select Plot Sample Histogram from the QA/QC section of the Illumina BeadArray Methylation workflow to bring up a Histogram tab

Again, no sample in the tutorial data set stands out (Figure 34).

 

Numbered figure captions
SubtitleTextSample histogram. Each sample is a line, M-values are on the horizontal axis and their frequencies on the vertical axis. Two peaks correspond to two probe types (I and II) present on the MethylationEPIC array. Sample colors correspond to a categorical attribute (i.e. HPSCState)
AnchorNamesample histogram

Image RemovedImage Added

 

 

Section Heading

Section headings should use level 2 heading, while the content of the section should use paragraph (which is the default). You can choose the style in the first dropdown in toolbar.

 

 

Page Turner
button-linkstrue

 

Additional assistance

 

Rate Macro
allowUsersfalse