Dr. Oz recently did a show on genetics and weight loss--nutrigenomics, which was a combination of really good information and some not so good information.
There is something to the idea that your genetic make-up has a lot to do with what kind of diet you will respond to and how. While isolating the particular genes that identify which diet is best for you is somewhat new, the general hypothesis has been around for awhile in the form of both anthropological and ethnographic nutrition. Which is to say: we've often looked at our human ancestors in general, and our specific ancestors and their places of origin in order to determine what diets suit which people best.
This is the foundation for the paleolithic diet--eat only what the cave men ate, nuts, berries, leafy greens, and meat--and for the real food movement, which basically says, if it has a food label, you shouldn't eat it.
The nutrigenomic idea is slightly misleading--the report suggests you have a gene that tells us exactly what you should eat. This is not exactly the case. You have genes that tell us what sort of disease processes and metabolic tendencies that you've inherited. What diet that means you should have is, at best, a hypothesis, (what I like to call a SWAG, a scientific, wild-ass guess) much of it based on faulty premises, like the idea that low-fat dieting has a positive effect on stroke and heart-attack, or the idea of a calorie and its relationship to nutrition and obesity.
For that is the most potentially misleading part of Dr. Oz's report: the idea that the bottom-line is "calories in, calories out"--the laws of thermodynamics tell us that this is not the case. It takes more energy to process a protein "calorie" than a carbohydrate "calorie". I use calorie in quotes, because even the concept of a calorie is rather bizarre and very flawed.
It's not just calories in, calories out. Not by a long shot.