. 2021 Aug 17;25(3):141–149. doi: 10.4235/agmr.21.0080

Table 1.

Top 10 DNA methylation sites associated with the prediction of age, frailty index, or a sequence of randomly generated numbers

Age^a)		FI34^b)		Random number^c)
CpG site^d)	Importance^e)	CpG site^d)	Importance^e)	CpG site^d)	Importance^e)
cg18404925	30.4	cg10458216	12.9	cg02271560	12.0
cg21070864	27.0	cg26188685	4.6	cg10332616	7.9
cg26811352	14.9	cg01290856	4.4	cg26565544	4.9
cg16624788	14.5	cg23765086	4.2	cg00981070	4.7
cg01290856	12.7	cg11177404	4.1	cg12645852	4.3
cg13480818	10.9	cg02716646	3.9	cg04365609	3.7
cg03767951	9.7	cg10043954	3.9	cg05461182	3.7
cg01874084	9.5	cg21561157	3.8	cg05935052	3.6
cg07935657	7.3	cg10223982	3.2	cg03475190	3.5
cg13567404	7.2	cg05304366	3.2	cg10697491	3.4

^a)

The calendar ages (60–103 years) of 211 participants in the Louisiana Healthy Aging Study⁷²) were regressed on DNA methylation values (beta) of 10,000 randomly selected CpG sites using a random forest regression method in R.

^b)

The same random forest as in ^a), with a frailty index, FI34, instead of calendar age.

^c)

The same random forest as in ^a), with a variable of randomly generated numbers instead of calendar age.

^d)

Probes used to assess DNA methylation levels of the CpG sites using the Infinium HumanMethylation450K BeadChip Kit (Illumina Inc., San Diego, CA, USA).

^e)

Scaled permutation-based variable importance scores calculated using the random forest.