Methylation Browser



























































linear model methylation vs covariate:

              

            

Sex table:


              

Site table:


              

Age table:


              

Stromal content table:


              
Categorized splitting by the median:

              

Tumoral variables tables:

CIMP table:

              
Mutations table:

              
CNV table:

              
CNV categorized table (splitting by median):

            

PLOTS PANEL

  • Zoom: slider to select the region to explore.

  • Methylation plot: shows the methylation profile for each probe and for each covariate value.

  • Methylation Difference plot: shows the difference in methylation median for each CpG. Differences are displayed as arrows taking a value of the covariate as baseline. Differentially methylated probes are indicated with an asterisk on the top of the plot.

  • CpG islands: shows CpG islands, CpG island shores, CpG island shelves and open sea.

  • Genes: shows all genes that are inside the selected region.

  • Expression vs Methylation plot:
    • if covariate is tissue:
      - a first plot of the selected probe methylation (beta values) vs the selected gene expression (rescaled PCA first component) is shown. Colour represent the different tissues. The global correlation coefficient is shown in the title and the correlation coefficient for each tissue is shown in the legend. Paired samples are joined by a gray line.
      - a second plot of the difference between paired adjacent and tumoral samples is displayed if both tissues are selected. The correlation coefficient is shown in the title.

    • if covariate is not tissue:
      - a first plot of the selected probe methylation (beta values) vs the selected gene expression (rescaled PCA first component) is shown. Colour represent the levels of the selected covariate. The global correlation coefficient is shown in the title and the correlation coefficient for each covariate level is shown in the legend. Paired samples are joined by a gray line.
      - a second plot of the difference between paired adjacent and tumoral samples is displayed if both tissues are selected. Colour represent the levels of the selected covariate in tumoral samples. The correlation coefficient is shown in the title and the correlation coefficient for each covariate level is shown in the legend.
      - a last plot representing the selected probe methylation (beta values) vs the selected gene expression (rescaled PCA first component) is displayed for each covariate value. Colour represent the different tissues. The global correlation coefficient is shown in the title and the correlation coefficient for each tissue is shown in the legend. Paired samples are joined by a gray line.

    • A table with the statistical analysis for the selected probe as a subset of the table in the table panel.


TABLE PANEL

A table for each CpG in the selected region is displayed.
For each value of the chosen covariate the number of patients, the mean and standard deviation of beta values are shown.
To compare between values of the covariate, the table includes the difference in beta value, the standard error, the statistic and the p-value.
The adjusted p-value taking into account all CpGs in the region is also included.



SAMPLE DESCRIPTIVE PANEL

A table for each covariate displaying the descriptive statistics for each tissue.



USAGE AND PARAMETERS

  1. Enter a gene name.

  2. Covariates
    • Choose a covariate to perform the methylation analysis (default = tissue). Numerical covariates are cuted by their median value.

    • Filter the values of the different available covariates.

  3. Methylation plot type:
    • Boxplot: boxplot for each CpG and for each covariate value (default).

    • Medians: median for each CpG and for each covariate value.

    • Stripchart: values for each CpG, for each patient and for each covariate value.

  4. Statistical analysis:
    • Statistical test:
      • t-test (beta): classical t-test (default).

      • t-test logit (beta): classical t-test using logit transformation of the beta values log(betas/(1-betas)).

      • non-par: classical t-test using ranks of the beta values.

    • P-value adjustment method: The available methods to adjust the p-value for multiple comparisons of CpGs are Bonferroni (default) and FDR.

    • Significance level: Threshold to show a CpG as differentially methylated in the methylation difference plot.

  5. Display plots:
    • Show CpG islands

    • Show genes

    • Display Selected Gene Isoforms

  6. Correlation between expression and methylation:
    • Select a probe: (empty by default) Choose a probe to display the correlation between expression and methylation. It is also possible and equivalent to click on a probe in the first or second plot.

    • Correlation method: The available methods are Pearson's r (default), Kendall's tau and Spearman's rho.

    • Display regression line


  7. Zoom: Adjust the region to explore using the sliding bar on the top of the plot panel. By default it is set to the selected gene range but the user can zoom in to see in more detail a part of the gene or zoom out a maximum of 100Kb out of the selected gene.