Abstract: Disclosed herein are methods for detecting differentially methylated CpG islands associated with epigenetic changes in a subject using discrete cosine transform (DCT) and spectral clustering on non-native data. Based, in part, on algorithms continually improved through machine learning and the application of a combination of DCT and spectral clustering when applying a deep learning program on non-native data, the methods and systems generate a customized report on an individual's overall health. This contains a neural network-trained assessment of the individual's genome, including differences in DNA methylation and gene expression—quantitative results which can predict the onset of developing health concerns and conditions associated with aging.

Abstract: Disclosed herein are methods for detecting differentially methylated CpG islands associated with epigenetic changes in a subject. Based, in part, on algorithms continually improved through machine learning, the methods and systems generate a customized report on an individual's overall health. This contains a neural network-trained assessment of the individual's genome, including differences in DNA methylation and gene expression—quantitative results which can predict the onset of developing health concerns and conditions associated with aging. The results can then be privately and conveniently accessed and shared with health providers to deliver a qualitative assessment backed by clinical judgment, and to facilitate deploying targeted treatments of identified conditions associated with aging.