Therapeutic approaches to treating obesity based on dietary modification and exercise have been, in general, unsuccessful. Pharmacotherapies for obesity, when coupled with lifestyle modification, can enhance weight loss, however there is inter-individual variation in treatment response. An alternative strategy to a ‘one-size-fits-all’ approach to obesity management is ‘Personalised Medicine’, in which information about a person’s genes, environment and other characteristics are used to develop an individualized treatment plan to optimize results. To date, clinical trials have focused exclusively on genotype-tailored interventions for weight loss with varying success and results remain unclear1,2. Similarly, pharmacogenomics studies have been unable to accurately predict responders and non-responders to obesity pharmacotherapies3. This highlights the need for a whole systems approach in obesity, in which genomic analysis is combined with analysis of dynamic variables such as the proteome, metabolome and microbiome, thereby integrating both genetic background and response to the environment. A recent study by Zeevi et al. reveals that extensive monitoring of a human cohort for variations in dietary intake, lifestyle, host phenotype, and the gut microbiome enabled the development of a machine-learning algorithm that accurately predicted the individual glycemic response to meals. This provides evidence that personalized intervention in obesity may be possible4.