Using composite phenotypes to reveal hidden physiological heterogeneity and model oxygen saturation variation of high altitude acclimatization in a Chinese Han longitudinal cohort

Altitude acclimatization is the physiological process of the human body adjusting to the decreased availability of oxygen. Since several physiological processes are involved and the relation among them is complicated, analyses of single-traits is insufficient in revealing the complex mechanism of altitude acclimatization. In this study, we examined whether these physiological responses could be studied as composite phenotypes which are represented by a linear combination of physiological traits. We developed a strategy which combines both spectral clustering and PLSPM to define composite phenotypes. We captured 14 composite phenotypes from 28 physiological traits of altitude acclimatization. Using these composite phenotypes, we applied k-means to reveal hidden physiological heterogeneity in altitude acclimatization. Furthermore, we employed linear regression to systematically model oxygen saturation (SpO2) changes in altitude acclimatization and evaluated the model fitness performance. And composite phenotypes based Model 2 has better fitness than single-traits based Model 1 in all measurement indices. Therefore, this new strategy of defining and applying composite phenotypes can be considered as a general strategy of complex traits.