# Methylation array analysis ## 1.What does the letter-number combination in the coluhttps://hackmd.io/mn Array stand for? Row number, Column number of the sample, likely the position on the well plate the samples where prepared in -> good to be able to back-trace problems with the sample ## 2. What are likely the most interesting columns for our analyses? Chr, pos, name, probe, UCSC gene name ## 3. How is the quality of the samples? Can we use all samples in the following analyses, or should we exclude samples? All samples but one show good overall quality. Remove the birth sample ## 4. How many samples are left after the removal of poor quality data? Re-plot the data without the poor data sample. 10 ## 5. Try out the plotSex function. What does it show? All males in the sample set ## 6. How do we know that the normalization was successful? Because after normalization the lines of the samples are equal ## 7. What sample information is used to colour code the samples in the plots? What do these plots tell us about the samples and what we can expect in the following analyses? Sample identity and treatment typ. A lot of variation between the samples, some variation between the treatments. ## 8. Re-plot the MDS plots with the filtered data. How has the positioning of the data points changed? What does this tell us for our subsequent analyses? The between indivudual differences have been leveled a bit and the treatment differences are now more important. However, the individual should still be taken into account when doing the analyses and should be corrected for. ## 9. What is the meaning of the peaks in the plots? How do the ranges of beta or M values make them easier to interpret/ easier to handle for statistical purposes? Low value peak=unmethylated High value peak=methylated beta 0 and 1, M neg and pos ## 10. Can you retreive the top table for another contrast? ## 11. What can you see in the plots? Does what you see make sense in the light of our analyses? ## 12. Look at results.ranges. What information can we find in there? ## 13. Look at the plot. What can you see? How can you change the code to inspect other differentially methylated regions?