Steve
Horvath (shorvath at mednet.ucla.edu)
|
|
Using 82 Illumina DNA methylation array data sets (n=7844)
involving 51 healthy tissues and cell types, I developed a multi-tissue predictor
of age which allows one to estimate the DNA methylation (DNAm) age of most
tissues and cell types. DNAm age has the following properties: a) it is close
to zero for embryonic and induced pluripotent stem (iPS) cells, b) it
correlates with cell passage number, c) it gives rise to a highly heritable
measure of age acceleration, and d) it is applicable to chimpanzee tissues.
http:/genomebiology.com/2013/14/10/R115
Frequently asked questions (FAQs): faq.htm
Updated results:
1) cancer tissue: Oct 27, 2013
2) Figure 3E correction2
DNAm age calculator webpage (below) contains
information on how to calculate DNA methylation (DNAm) age based on data
measured using the Illumina Infinium platform (e.g. 450K or 27K data).
The age calculator automatically outputs the
estimated DNAm age. After uploading the data, the function will return an Excel
file whose rows report the estimated DNAm age of each subject and additional
information.
Access Online Age Calculator: https://dnamage.genetics.ucla.edu/
Oct 15, 2014. There has been a major update.
Now, the epigenetic clock software allows one to select an advanced analysis
for blood tissue (measured on the Illumina 450K array).
Example data based
on the Illumina DNA Infinium platform
To evaluate the advanced analysis in blood, upload the following zipped file and corresponding sample annotation file.
datSampleBloodIllumina450K.csv
A tutorial that describes how to calculate
DNAm age based on Illumina DNA Infinium 27K or 450K data is available here:
TUTORIALonlineCalculator.docx or TUTORIALonlineCalculator.pdf
Save these files into the directory where your DNA meth data are located.
To run this tutorial, also download the following example data sets:
Example 55
Example: Blood
Illumina 450K
To evaluate the advanced analysis in blood, upload the following zipped file and corresponding sample annotation file.
datSampleBloodIllumina450K.csv
Requirements
To run the advanced analysis in blood, your methylation data need to contain the CpGs listed in the file
Further, I strongly recommend you upload a sample annotation file that contains the variables "Age", "Female", "Tissue" (note the capitalization/spelling).
Instructions on how to upload large data sets can be found in TUTORIALonlineCalculator.docx or TUTORIALonlineCalculator.pdf.
The advanced analysis in blood is not yet
available as R script. However, the original standard analysis is implemented
in
the following R software tutorial that shows
how to calculate DNAm age based on Illumina DNA Infinium 27K or 450K data is
available. These R scripts carry out normalization steps and show how to
estimate DNAm age.
R
Tutorial 1: Analysis of data set 55.
Word document: TUTORIAL1.docx
PDF file: TUTORIAL1.pdf
1) probeAnnotation21kdatMethUsed.csv
Save these files into the directory where your DNA meth data are located.
To run this tutorial, also download the following example data sets:
6) MethylationDataExample55.csv
7) SampleAnnotationExample55.csv
1) Oct 14, 2014. The epigenetic clock was used
to demonstrate that "Obesity
accelerates epigenetic aging of human liver"
Horvath S, Erhart W, Brosch M, Ammerpohl O, von Schönfels W, Ahrens M, Heits N, Bell JT, Tsai PC, Spector TD, Deloukas P, Siebert R, Sipos B, Becker T, Röcken C, Schafmayer C, Hampe J (2014) Obesity accelerates epigenetic aging of human liver. Proc Natl Acad Sci U S A. pii: 201412759. PMID: 25313081
2) Oct 15, 2014. Major software update. Te epigenetic clock software allows one to select an advanced analysis for blood tissue (measured on the Illumina 450K array). Results: various measures of age acceleration and estimates of cell counts. See the online tutorial for details.
Please email Steve Horvath with questions (shorvath at mednet.ucla.edu), subject heading: DNAmAge.
NIH 5R01AG042511-02
SH gratefully acknowledges the many researchers who made their DNA-methylation datasets publically available and responded to his email requests. Special thanks to Gene Expression Omnibus, ArrayExpress, and The Cancer Genome Atlas.