cpg methylation by infinium human methylation 450 k beadchip arrays 450 k  (INFINIUM Inc)

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: a , Public human blood DNA methylation datasets and age distributions. b , Youthful methylation level and age-related drift of CpGs inside and outside of repetitive elements. c , Methylation drift rate of CpGs grouped according to major repeat class. Selfish (transposons) and nonselfish repeats were grouped separately. Only RE classes represented by 100 or more CpGs in the Infinium array are shown. d , Methylation drift rate of CpG in L1s, grouped according to family and sorted according to average sequence length, a proxy of evolutionary age. e , Methylation drift rate of CpG in LTR retrotransposons, grouped according to family and sorted according to average sequence length. d , e , Only families represented by 40 or more CpGs in the Infinium array are shown. c – e , Boxes show the median, and the 25th and 75th percentiles. The whiskers extend to the 25th and 75th percentiles ± 1.5 times the interquartile range (IQR). Points outside the whiskers are not shown.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: DNA Methylation Assay, Methylation, Sequencing

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: Consensus position of CpGs and methylation drift rate of selected L1s and LTR retrotransposons.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Methylation

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: ( A ) Methylation drift rate of CpGs in SINEs, grouped by family. ( B ) Methylation drift rate of CpGs in SVAs, grouped by family. ( C ) Methylation drift rate of CpGs in DNA transposons, grouped by family. ( D , E , F ) Age coefficient of methylation at SINE, SVA and DNA transposon CpGs after adjustment for CpG density and youthful methylation level. ( A - F ) Only families represented by 40 or more CpGs in the Infinium array were shown. Families are sorted by average sequence length. Boxes show the 25th, median, and 75th percentiles. Whiskers extend to the 25th/75th percentiles ± 1.5*IQR. Points outside of whiskers not shown.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Methylation, Sequencing

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: a , Trends of methylation drift based on youthful methylation levels. b , Trends of methylation drift based on local CpG density. c , d , Age coefficient of methylation at LINEs ( c ) and LTR retrotransposons ( d ) after adjustment for CpG density and youthful methylation level. Only families represented by 40 or more CpGs in the Infinium array are shown. e , TF binding motifs enriched at young L1s and associated with increased or decreased methylation drift rate in young L1s (two-sided Wilcoxon rank-sum test, P < 0.05 for both adjusted and unadjusted age drift rate coefficient). P values are only shown for comparisons in young L1s; the exact P values are, in order, 0.021, 0.039, 0.028, 0.016. n of CpGs in old L1s: 341; n CpGs in young L1s: 621, of which 416 are flanked by an ARNTL motif, 589 by an NFKB2 motif, 375 by a FOXO1 motif and 390 by a HIC2 motif. Additional motifs are shown in Extended Data Fig. . c – e , Boxes show the median and the 25th and 75th percentiles. The whiskers extend to the 25th and 75th percentiles ± 1.5 times the IQR. Points outside the whiskers are not shown. * P < 0.05.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Methylation, Binding Assay

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: ( A ) Additional transcription factor binding motifs enriched at young L1s and associated with increased or decreased methylation drift rate within young L1s (Two-sided Wilcoxon rank-sum test p < 0.05 for both adjusted and unadjusted drift rate coefficient). P-values are only shown for comparisons within yL1s, exact P-values are 0.0029, 0.003, 0.038, 0.046 in order. N CpGs within old L1s: 341, N CpGs within young L1s: 621, of which 515 are flanked by an AR motif, 562 by an NR1I2 motif, 499 by an RXRA::VDR motif, and 454 by an SP5 motif. Boxes show the 25th, median, and 75th percentiles. Whiskers extend to the 25th/75th percentiles ± 1.5*IQR. Points outside of whiskers not shown. ( B ) Motifs enriched near L1PA2 CpGs with faster-than-median methylation drift vs those with slower-than-median drift. TFs that have similar motifs, or are part of the same TF family are grouped together. ( C ) Motifs enriched near L1HS CpGs with faster-than-median methylation drift vs those with slower-than-median drift. ( D ) Motifs enriched near L1PA4 CpGs with faster-than-median methylation drift vs those with slower-than-median drift.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Binding Assay, Methylation

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: a , Feature construction strategy. b , Test set performance of the individual CpG clock. c , Test set performance of the combined CpG clock. d , Benchmark of individual and combined CpG clocks against state-of-the art methylation clocks. The benchmark was performed on GSE64495 , which was not included in the training set of any of the clocks shown. e , Performance of a combined CpG clock trained on multi-tissue mouse RRBS data, tested using nested cross-validation. f , Age prediction on long-lived mouse strains compared to matching controls (two-sided Wilcoxon rank-sum test: * P = 0.026 and ** P = 0.002, respectively). Boxes show the median and 25th and 75th percentiles. The whiskers extend to the 25th and 75th percentiles ± 1.5 times the IQR. Points outside whiskers (outliers) are shown individually.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Methylation, Biomarker Discovery

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: a , Association between age acceleration and risk of cancer, CHD and mortality according to the individual and combined CpG clocks in the WHI BA23 dataset. Results were benchmarked against state-of-the-art chronological age (Horvath Pan-tissue and Horvath Skin & Blood) and biological age (Levine PhenoAge) clocks. Associations between predicted age acceleration and risks were determined with a Cox proportional hazards model, accounting for age. The bars represent coefficients ± s.e. b , Age trajectory of methylation at young (first row) and old L1s with the largest coefficients in the combined CpG clock (second row). Data are from GSE40279 . The orange dashed line shows a linear fit ± 95% confidence interval (CI), excluding patients older than 65 years. The teal line shows a locally estimated scatterplot smoothing fit ± 95% CI on the full age range.** P < 0.01, *** P < 0.001. c , Effect of cancer within 3 years and age on methylation of young and old L1s in the WHI data. Coefficients and P values were determined by fitting a linear model with the following formula: methylation ~ age + any cancer in 3 years. d , Performance of predictors of risk of cancer, CHD and mortality within 3 years, based on young L1 CpGs. The bars represent the mean area under the receiver operating characteristic curve (ROC a.u.c.) ± s.d. in tenfold cross-validation. Best and parsimonious models are shown.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Methylation, Biomarker Discovery

Journal: Nature Aging

Article Title: Transposable element 5mC methylation state of blood cells predicts age and disease

doi: 10.1038/s43587-024-00757-2

Figure Lengend Snippet: ( A ) Young L1 methylation in patients who developed cancer withing 3 y of sampling compared to healthy individuals, broken down by cancer site. Dots represent individual (future) cancer patients, and the mean is represented by a red triangle. The blue dashed line represents the average methylation in healthy individuals, while the orange dashed line represents average methylation over patients diagnosed with any cancer within 3 years. ( B ) Additional cancer, CHD and mortality predictors, trained as those in Fig. D but allowing THE1A and THE1C CpGs the feature selection process in addition to young L1 CpGs. Bars represent mean ROC-AUC + /- SD in 10-fold cross-validation. Best and parsimonious models are shown.

Article Snippet: Associations between age and Infinium array CpG methylation were determined using limma , with the age + sex + study design.

Techniques: Methylation, Sampling, Selection, Biomarker Discovery