Identification of type 2 diabetes subgroups through topological analysis of patient similarity

See allHide authors and affiliations

Science Translational Medicine  28 Oct 2015:
Vol. 7, Issue 311, pp. 311ra174
DOI: 10.1126/scitranslmed.aaa9364

Networks work for diabetes

Big problems require big solutions, and for complex diseases such as cancer or diabetes, the big solution is big data. One long-term goal of U.S. President Barack Obama’s Precision Medicine Initiative is to assemble medical and genetic data from at least one million volunteers. But how might researchers use all those data? Li et al. provide one answer by using patient electronic medical records (EMRs) and genotype data from Mount Sinai Medical Center in New York to characterize new subtypes of type 2 diabetes (T2D).

The group first clustered EMR data to identify T2D patients within the larger group. Topological analysis of the T2D group identified three new T2D subtypes on the basis of distinct patterns of clinical characteristics and disease comorbidities. Genetic association analysis identified more than 300 single nucleotide polymorphisms (SNPs) specific to each subtype. The authors found that classical T2D features such as obesity, high blood sugar, kidney disease, and eye disease, were limited to subtype 1, whereas other comorbidities such as cancer and neurological diseases were specific to subtypes 2 and 3, respectively. These distinctions might call for tailored treatment regimens rather than a one-size-fits-all approach for T2D. Although a larger sample size is needed to determine causal relationships, this study demonstrates the potential of precision medicine.

View Full Text

Stay Connected to Science Translational Medicine