Search:

PERSONALIZED MEDICINE

What we provide. We analyze data from whole-genome sequencing and whole-genome genotyping. We incorporate clinically realistic indicators of the health state of an individual, using epigenetic and genome-wide deletion and duplication analyses, together with cell biological and clinically relevant phenotypic parameters. We do not offer direct-to-consumer genomic tests.

Personalized genomics. Personalized genomics is undoubtedly going to be one of the most important areas of human medicine, although it is currently experiencing credibility issues as regards clinical relevance. This is a result of the large number of companies already offering diagnostic health tests that are best categorized as recreational genomics.

Solid whole genome data at the most basic DNA level are, however, now coming on stream. The recent publications of the near complete diploid genomes of J Craig Venter and James Watson have, nevertheless, led to extensive questioning of what individual genomes may reveal about one's future health and the clinical consequences of the medicines that an individual may take during periods of illness. For example, the cytochrome P450 drug metabolizing genes have been highlighted in this regard, as they are involved in metabolizing antidepressants, codeine, statins, warfarin, beta-blockers and antipsychotics. The detailed genomic differences in drug metabolizing polymorphisms between Venter and Watson in six of these genes have recently been published in Clinical Pharmacology and Therapeutics by the Venter laboratory, but claims about their clinical value have been tentative. Why is this the case?

George Gabor Miklos

Our analyses. Although the base level DNA data are exquisite, it is still clinically hazardous to make a decision about a patient's ability to metabolize drug combinations solely on the basis of their SNP content. Such decisions require additional inputs which must take into account lifestyle factors such as smoking and metabolic snapshots at the time of diagnostic testing. Even if an individual has a completely “normal” set of drug-metabolizing alleles, they can still be compromised by epigenetic silencing of gene activity which reduces their activities to zero.

Epigenetic silencing of active genes, or epigenetic reactivation of dormant genes, in particular organs and tissues requires genome-wide methylation analysis or data on post-translational histone modifications; these are not discernible from SNP analyses. In addition, it is only now being generally realised that there are thousands of duplications and deficiencies in normal human genomes. Their effects on drug metabolism can be substantial, depending on the particular combination of duplications and deficiencies in that individual and the penetrance, expressivity and altered epistatic interactions that culminate in altered fluxes through their networks in different tissues.   

A seemingly healthy individual who arrives at their doctor's office with a CD containing their entire 6 billion base pair diploid genome and expects to receive an estimate of future healthcare problems such as Alzheimer's, dementia or predisposition to cancer, is in a completely different category to a patient requiring emergency attention for acute renal failure. While physicians are experts at handling immediate health issues, very few physicians, counsellors or genetic testing companies can make robust clinical predictions about future health states on the basis of standard mono-level DNA sequence data. The major hurdle to assessing risk from the DNA level is that gene variants only predispose to a disease state in particular genetic and epigenetic backgrounds. Importantly, epigenomic contributions may largely overwhelm an inherited condition. It is a common finding that the same mutation in two different individuals may be associated with a disease in one person but not in the other. In addition, even when a deleterious genetic variant is present in a genome, it may be epigenetically silenced and hence of little clinical significance.

Since all human conditions are a combination of genetics and epigenetics, (including diet and lifestyle), estimating the risk for a future health state which may or may not arise in the lifetime of an individual is not the same as knowing the risk within a population.

The way ahead. We analyze genetic predispositions using multilevel criteria with special emphasis on epigenomic status modified by lifestyle, diet and use of medications.

 

2008, Ng, P. C. et al., Clinical Pharmacology and Therapeutics, 84, 306-309. Individual genomes instead of race for personalized medicine.