Omic personality: implications of stable transcript and methylation profiles for personalized medicine

Abstract

BackgroundPersonalized medicine is predicated on the notion that individual biochemical and genomic profiles are relatively constant in times of good health and to some extent predictive of disease or therapeutic response. We report a pilot study quantifying gene expression and methylation profile consistency over time, addressing the reasons for individual uniqueness, and its relation to N = 1 phenotypes.MethodsWhole blood samples from four African American women, four Caucasian women, and four Caucasian men drawn from the Atlanta Center for Health Discovery and Well Being study at three successive 6-month intervals were profiled by RNA-Seq, miRNA-Seq, and Illumina Methylation 450 K arrays. Standard regression approaches were used to evaluate the proportion of variance for each type of omic measure among individuals, and to quantify correlations among measures and with clinical attributes related to wellness.ResultsLongitudinal omic profiles were in general highly consistent over time, with an average of 67 % variance in transcript abundance, 42 % in CpG methylation level (but 88 % for the most differentiated CpG per gene), and 50 % in miRNA abundance among individuals, which are all comparable to 74 % variance among individuals for 74 clinical traits. One third of the variance could be attributed to differential blood cell type abundance, which was also fairly stable over time, and a lesser amount to expression quantitative trait loci (eQTL) effects. Seven conserved axes of covariance that capture diverse aspects of immune function explained over half of the variance. These axes also explained a considerable proportion of individually extreme transcript abundance, namely approximately 100 genes that were significantly up-regulated or down-regulated in each person and were in some cases enriched for relevant gene activities that plausibly associate with clinical attributes. A similar fraction of genes had individually divergent methylation levels, but these did not overlap with the transcripts, and fewer than 20 % of genes had significantly correlated methylation and gene expression.ConclusionsPeople express an “omic personality” consisting of peripheral blood transcriptional and epigenetic profiles that are constant over the course of a year and reflect various types of immune activity. Baseline genomic profiles can provide a window into the molecular basis of traits that might be useful for explaining medical conditions or guiding personalized health decisions.Electronic supplementary materialThe online version of this article (doi:10.1186/s13073-015-0209-4) contains supplementary material, which is available to authorized users.

Individualized Transcriptional Resolution of Complicated Malaria in a Colombian Study

“…Such individualization is reminiscent of our findings in healthy individuals [ 36 ]. In that study, we found, by contrasting the RNAseq profiles of a dozen of individuals at three six-month intervals, that all subjects consistently clustered together, whether considering the major PC of the entire transcriptome or the Axis scores.…”

“…In that study, we found, by contrasting the RNAseq profiles of a dozen of individuals at three six-month intervals, that all subjects consistently clustered together, whether considering the major PC of the entire transcriptome or the Axis scores. This gave rise to the notion of an “omic personality”, a consistent profile of gene expression over time that reflects the baseline healthy status of an individual [ 36 ]. This baseline expression is quite likely to have a major influence on the recovery profiles.…”

Disease-specific regulation of gene expression in a comparative analysis of juvenile idiopathic arthritis and inflammatory bowel disease

“…Blood Transcript Module and BIT axis analyses, both based on comprehensive analysis of existing whole blood gene expression datasets, confirm that these types broadly reflect differences in gene activity in the major immune sub-types, partly reflecting cell abundance, but also innate states of activity of biosynthetic, cell cycle, and cytokine signaling. Immunoprofiling by flow cytometry has established that individuals have baseline profiles, or omic personalities [ 71 ], to which they return after immunological perturbation but which are also influenced by such environmental factors as child-rearing [ 72 ]. Sub-type-specific blood gene expression should be seen in light of this immunological elasticity, as the heterogeneity among subjects may be more meaningful for disease risk than individual eQTL effects.…”

ACE2 expression in adipose tissue is associated with COVID-19 cardio-metabolic risk factors and cell type composition

“…To estimate the proportions of adipose and blood cell types in the adipose biopsies from the FUSION and METSIM studies, a reference transcriptome was created using RNA-Seq of whole blood (GEO accession GSE67488 54 ), and of four types of purified adipose cells (adipocytes, macrophages, CD4+ T cells, and microvascular endothelial cells) described in Glastonbury et al 33 . For the FUSION study, reference transcriptome reads were aligned to the hg19 reference genome using the same read mapping and quality control procedure as used for the FUSION adipose RNA-Seq data 27 .…”