Interpreting TCGA 450K data

Question

[ChIP]

Hello Everyone,

I was having a look at TCGA 450K data and I downloaded also the data at two levels, ie., level 2 and level3.

Level 2, looks like this:

Composite Element REF    Methylated_Intensity    Unmethylated_Intensity    Detection_P_value
cg00000029    2488.00579881129    2281.3142892634    0
cg00000108    8943.62421381116    336.745081332759    0
cg00000109    3827.0493383932    219.47270455192    0
cg00000165    263.820225926362    2355.4623873349    0
cg00000236    3733.92206994152    722.124674419151    0
cg00000289    1503.38837282518    277.513187258243    0
cg00000292    9650.668135454    1618.68980927383    0
cg00000321    5918.34421148769    4384.53097589033    0

while level 3, looks like this:

Composite Element REF    Beta_value    Gene_Symbol    Chromosome    Genomic_Coordinate
cg00000029    0.521668865344633    RBL2    16    53468112
cg00000108    NA    C3orf35    3    37459206
cg00000109    NA    FNDC3B    3    171916037
cg00000165    0.100722321673368        1    91194674
cg00000236    0.837944995677383    VDAC3    8    42263294
cg00000289    0.844172640712807    ACTN1    14    69341139
cg00000292    0.856363617411664    ATP2A1    16    28890100
cg00000321    0.574436174742582    SFRP1    8    41167802

Anyone, who has used this data can please tell me how I can Identify the regions/genes that are hyper and hypo-methylated.

Thank you

Answer 1

[Sean Davis]

What you have downloaded are per-sample files describing the methylation values at single CG sites. Hyper- and hypo-methylation are "relative" terms, so you will need to define two (or more) groups and then look for differential methylation.

[ChIP]

Hi Sean,

Thank you for your answer but I have another one for example say I have another group of values for second sample and I have to compare them and then identify regions that are hyper or hypo methylated? then which level of data shall I use (level2 or level 3) and how can I use it ? and secondly is it possible to use 3 Vs 13 samples to identify hyper and hypomethylated regions.

Thank you

[jianggl2000]

You can use the level3 data, and compare the beta values between 2 or more groups. Logit transformation to M values will help with the statistical analysis.

You can also summarize multiple probes annotated to the same CpG island to a single value (e.g. mean) and compare the difference at the island level. Also M value will be helpful in comparison.

3 vs. 13 is OK.