DNA methylation test

Questions about Horvath's DNAm Age

    Steve Horvath's DNA methylation clock is based on the presence or absence of methylation at 353 (out of many thousand) CpG sites on a human or animal genome.  The sites selected were best correlated with age while being independent of cell type.  This clock has become the standard for determining the "biological age" of tissue, blood, and urine samples. I've had my DNAm age done by Zymo on two blood samples taken at the same time using some version of this technique.  They were both just a bit below my calendar age and one was about 2 years different from the other.

     I have recently been studying several papers by Horvath and others on DNA methylation vs. aging, and I've also watched his video interviews and presentations.  However, as a physicist interested in mechanism, I find that there are basic questions that do not seem to be addressed in any of these DNAm papers or videos that I have read or watched.  Perhaps some of you are more knowledgeable than am I about this area, so let me present my questions here: 

(1) Of the 353 CpGs selected by Horvath for use in determining the DNAm age, how many of them have a positive correlation (increase with age) and how many of them have a negative correlation (decrease with age)?

(2) Of the 353 CpGs selected for use in determining the DNAm age, how many of them are located in the non-coding region of DNA, how many of them are located in the coding region of DNA, and how many are located in CpG islands in protein promoter regions?

(3) Does the methylation of a particular DNA region cause gene expression or gene silencing?  What's the mechanism?

(4) David Sinclair discusses epigenetic acetylation and methylation almost interchangeably.  What are the respective roles of acetylation and methylation in determining the activation of a DNA region (switching on/off a gene)?  Is there also an acetylation clock?

(5) Is there any discernible pattern to the methylated DNA locations of the very young vs. the very old?

(6) If a subject's DNA was cleared of all methylation, what would be the result?

Answers, anyone?

A Few Answers to my DNAm Age Questions

    I've read further and have found tentative answers to some of my questions listed above.

(1)   About 2/3 of the CpG sites used by Horvath have methylation that is positively correlated with age, while about 1/3 have methylation that is negatively correlated with age.  That means (see below) that twice as many genes are being silenced as we age as are being newly expressed (at least for those sites that show up in the profile Horvath uses).

(2)  Many (most?) of the Horvath sites are in CpG islands in protein promoter regions.

(3&4)  Methylation tends to silence gene expression by blocking the promoter region, while acetylation tends to promote gene expression by exposing the promoter region to transcription factors. 

(6) The subject would die.  The methylation pattern is part of the selective epigenome that fixes the functions of particular cells (all with identical DNA) so that they become neurons or muscle cells or liver cells or skin cells or fat cells.   With no methylation they would lose their identity and cease to function correctly.