Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE credits
Epigenetic regulation modifies gene expression during the human life cycle and in response
To environmental factors, without changing the genetic code. Among epigenetic markers, DNA methylation is readily measurable in larger scale using the Illumina 450k microchip array, with around half a million interrogated sites per sample. Recently an age predictor, named ”theepigenetic clock”, based on CpG sites from this array was published, yielding surprisingly high correlation with age (R=0.95)
Here, previous Illumina 450k measurements of a longitudinal dataset consisting of Swedishtwins, including 1184 samples from six different measurement waves (time points), in total including 471 individuals, has been pre-processed and used to evaluate the epigenetic clock. Different normalization methods have been tested and compared for the dataset, and a pre-processing pipeline has been developed. Further, cross-sectional and total correlation for the epigenetic clock age predictor have veen determined. Age acceleration, previously associated with different age related phenotypes and defined as chronological age subtracted from presdicted age, has been calculated. Finally, longitudinal age differences for chronological and predicted age have been determined and compared.
The epigenetic clock has been shown to be strongly associated with age for the dataset (R=0.64), though not as strong as in the initial article. Mean age acceleration has been found to be negative for the whole dataset at 9.3 years (SD 8.7).
These results could point toward the predictor being badly adjusted for older ages, the dataset consisting of individuals with phenotapes associated with negative age acceleration, lacking sample quality control, or a combination of these. In short, these results further supportthe epigenetic clock a sa relatively reliable predictor of age, but with a lower correlation than previously reported.