No Average People

It's time to stop thinking about people as averages and start personalizing preventive medicine.

Full-text access for registered users only. Existing users login here.
New to GenomeWeb? Register here quickly for free access.

Where practicality, social

Where practicality, social acceptability (right or wrong), and scientific probabilities meet is now, has always been, and will always be the point of fierce battles. That is where one man's passion will be the opposite of another person's passion, and science should be the "passionless" participant. So, calm down, Koury and Topol

This is a specious arguement.

This is a specious arguement. You don't prescribe based on averages. You prescribe to the individual based on what you have observed in large patient cohorts. You can't build a risk prediction model based on a single patient observation.

We don't deal in averages. We deal in averages with uncertainies. Its just that most practitioners don't pay attention to the confidence intervals.

I agree with Dr. Khoury. In

I agree with Dr. Khoury. In order to prescribe treatments based on disease risk, it will be important to have risk factors that are highly informative for all individuals in a population. For example, whether or not an injured person receives blood transfusion, the risk is assessed by using markers such as A-B-O and RH blood biomarkers - something that does not just statistically enrich but prevents important known immunologic responses for the individual. Trying to do this transfusion screen with a single common blood antigen - for example, the A allele, enriches over no information at all, but is certainly not as informative as using the ABO genotyping for broadly matching blood types being transfused with the recipient patient - including haplotypes of decreased risk. Similar highly informative haplotyping of TOMM40-rs10523524 for risk of MCI/Alzheimer's disease is currently being validated in a large delay of onset [aka prevention] trial - the TOMMORROW Study. We are entering into an era where human genetic information in an individual can be widely assessed but needs to be tested for efficacy or adverse effects in regulated medical environments. Extremely highly variable, accurately measured, markers at a single locus, if validated in clinical trials, are the first real paradigm for predicting drug response at the level that does not call individuals "average" or "enriched" but comes with prospectively tested variations of highly informative human genotypes and haplotypes. [HLA B-5701 for an adverse event abacavir hypersensitivity] It becomes possible to map responses in populations where highly polymorphic markers at specific loci are tested in using phylogenetic mapping analyses. [EX.: Influenza mutations for annual update of vaccines.] None of us is average, plus or minus an acceptable mistake range, but it is possible to discover highly informative markers for measurable clinical responses during drug development - and also for using old drugs for new tricks.