Application of deep neural survival networks to the development of risk prediction models for diabetes mellitus, hypertension, and dyslipidemia.

OBJECTIVES: : Although numerous risk prediction models have been proposed, few such models have been developed using neural network-based survival analysis. We developed risk prediction models for three cardiovascular disease risk factors (diabetes mellitus, hypertension, and dyslipidemia) among a working-age population in Japan using DeepSurv, a deep feed-forward neural network. METHODS: : Data were obtained from the Japan Epidemiology Collaboration on Occupational Health Study. A total of 51 258, 44 197, and 31 452 individuals were included in the development of risk models for diabetes mellitus, hypertension, and dyslipidemia, respectively; two-thirds of whom were used to develop prediction models, and the rest were used to validate the models. We compared the performances of DeepSurv-based models with those of prediction models based on the Cox proportional hazards model. RESULTS: : The area under the receiver-operating characteristic curve was 0.878 [95% confidence interval (CI) = 0.864-0.892] for diabetes mellitus, 0.835 (95% CI = 0.826-0.845) for hypertension, and 0.826 (95% CI = 0.817-0.835) for dyslipidemia. Compared with the Cox proportional hazards-based models, the DeepSurv-based models had better reclassification performance [diabetes mellitus: net reclassification improvement (NRI) = 0.474, P  ≤ 0.001; hypertension: NRI = 0.194, P  ≤ 0.001; dyslipidemia: NRI = 0.397, P  ≤ 0.001] and discrimination performance [diabetes mellitus: integrated discrimination improvement (IDI) = 0.013, P  ≤ 0.001; hypertension: IDI = 0.007, P  ≤ 0.001; and dyslipidemia: IDI = 0.043, P  ≤ 0.001]. CONCLUSION: : This study suggests that DeepSurv has the potential to improve the performance of risk prediction models for cardiovascular disease risk factors.

Identification of dually acylated proteins from complementary DNA resources by cell-free and cellular metabolic labeling.

To establish a strategy to identify dually fatty acylated proteins from cDNA resources, seven N-myristoylated proteins with cysteine (Cys) residues within the 10 N-terminal residues were selected as potential candidates among 27 N-myristoylated proteins identified from a model human cDNA resource. Seven proteins C-terminally tagged with FLAG tag or EGFP were generated and their susceptibility to protein N-myristoylation and S-palmitoylation were evaluated by metabolic labeling with [(3)H]myristic acid or [(3)H]palmitic acid either in an insect cell-free protein synthesis system or in transfected mammalian cells. As a result, EEPD1, one of five proteins (RFTN1, EEPD1, GNAI1, PDE2A, RNF11) found to be dually acylated, was shown to be a novel dually fatty acylated protein. Metabolic labeling experiments using G2A and C7S mutants of EEPD1-EGFP revealed that the palmitoylation site of EEPD1 is Cys at position 7. Analysis of the intracellular localization of EEPD1 C-terminally tagged with FLAG tag or EGFP and its G2A and C7S mutants revealed that the dual acylation directs EEPD1 to localize to the plasma membrane. Thus, dually fatty acylated proteins can be identified from cDNA resources by cell-free and cellular metabolic labeling of N-myristoylated proteins with Cys residue(s) close to the N-myristoylated N-terminus.

Protein N-myristoylation is required for cellular morphological changes induced by two formin family proteins, FMNL2 and FMNL3.

The subcellular localization of 13 recently identified N-myristoylated proteins and the effects of overexpression of these proteins on cellular morphology were examined with the aim of understanding the physiological roles of the protein N-myristoylation that occurs on these proteins. Immunofluorescence staining of HEK293T cells transfected with cDNAs coding for the proteins revealed that most of them were associated with the plasma membrane or the membranes of intracellular compartments, and did not affect cellular morphology. However, two proteins, formin-like2 (FMNL2) and formin-like3 (FMNL3), both of them are members of the formin family of proteins, were associated mainly with the plasma membrane and induced significant cellular morphological changes. Inhibition of protein N-myristoylation by replacement of Gly2 with Ala or by the use of N-myristoylation inhibitor significantly inhibited membrane localization and the induction of cellular morphological changes, indicating that protein N-myristoylation plays critical roles in the cellular morphological changes induced by FMNL2 and FMNL3.