Prediction of gestational diabetes mellitus in Asian women using machine learning algorithms.

This study developed a machine learning algorithm to predict gestational diabetes mellitus (GDM) using retrospective data from 34,387 pregnancies in multi-centers of South Korea. Variables were collected at baseline, E0 (until 10 weeks' gestation), E1 (11-13 weeks' gestation) and M1 (14-24 weeks' gestation). The data set was randomly divided into training and test sets (7:3 ratio) to compare the performances of light gradient boosting machine (LGBM) and extreme gradient boosting (XGBoost) algorithms, with a full set of variables (original). A prediction model with the whole cohort achieved area under the receiver operating characteristics curve (AUC) and area under the precision-recall curve (AUPR) values of 0.711 and 0.246 at baseline, 0.720 and 0.256 at E0, 0.721 and 0.262 at E1, and 0.804 and 0.442 at M1, respectively. Then comparison of three models with different variable sets were performed: [a] variables from clinical guidelines; [b] selected variables from Shapley additive explanations (SHAP) values; and [c] Boruta algorithms. Based on model [c] with the least variables and similar or better performance than the other models, simple questionnaires were developed. The combined use of maternal factors and laboratory data could effectively predict individual risk of GDM using a machine learning model.

Fingerprint ridges allow primates to regulate grip.

Fingerprints are unique to primates and koalas but what advantages do these features of our hands and feet provide us compared with the smooth pads of carnivorans, e.g., feline or ursine species? It has been argued that the epidermal ridges on finger pads decrease friction when in contact with smooth surfaces, promote interlocking with rough surfaces, channel excess water, prevent blistering, and enhance tactile sensitivity. Here, we found that they were at the origin of a moisture-regulating mechanism, which ensures an optimal hydration of the keratin layer of the skin for maximizing the friction and reducing the probability of catastrophic slip due to the hydrodynamic formation of a fluid layer. When in contact with impermeable surfaces, the occlusion of the sweat from the pores in the ridges promotes plasticization of the skin, dramatically increasing friction. Occlusion and external moisture could cause an excess of water that would defeat the natural hydration balance. However, we have demonstrated using femtosecond laser-based polarization-tunable terahertz wave spectroscopic imaging and infrared optical coherence tomography that the moisture regulation may be explained by a combination of a microfluidic capillary evaporation mechanism and a sweat pore blocking mechanism. This results in maintaining an optimal amount of moisture in the furrows that maximizes the friction irrespective of whether a finger pad is initially wet or dry. Thus, abundant low-flow sweat glands and epidermal furrows have provided primates with the evolutionary advantage in dry and wet conditions of manipulative and locomotive abilities not available to other animals.

Algorithmic decomposition for efficient multiple nuclear spin detection in diamond.

Efficiently detecting and characterizing individual spins in solid-state hosts is an essential step to expand the fields of quantum sensing and quantum information processing. While selective detection and control of a few 13C nuclear spins in diamond have been demonstrated using the electron spin of nitrogen-vacancy (NV) centers, a reliable, efficient, and automatic characterization method is desired. Here, we develop an automated algorithmic method for decomposing spectral data to identify and characterize multiple nuclear spins in diamond. We demonstrate efficient nuclear spin identification and accurate reproduction of hyperfine interaction components for both virtual and experimental nuclear spectroscopy data. We conduct a systematic analysis of this methodology and discuss the range of hyperfine interaction components of each nuclear spin that the method can efficiently detect. The result demonstrates a systematic approach that automatically detects nuclear spins with the aid of computational methods, facilitating the future scalability of devices.

Metabolite profiling of enzymatically hydrolyzed and fermented forms of Opuntia ficus-indica and their effect on UVB-induced skin photoaging.

Fermentation of natural products is emerging as an important processing method and is attracting a lot of attention because it may have the advantage of having a new biological function. In this study, fruits of Opuntia ficus-indica were enzymatically hydrolyzed and then fermented with two species of yeast. We identified novel prominent markers in enzymatically hydrolyzed O. ficus-indica (EO) and fermented O. ficus-indica (FO) samples by using an ultra-performance liquid chromatography/quadrupole time-of-flight mass spectrometry. We also evaluated the effect of EO and FO on photoaging of skin cells exposed to ultraviolet radiation. We identified the major fermented metabolite in the FO as ferulic acid. Our in vitro study indicated that FO significantly enhanced the concentration of pro-collagen type 1 than the EO, by increasing the TGF-ß1 production.

Direct observation of the primary and secondary C-Br bond cleavages from the 1,2-dibromopropane photodissociation at 234 and 265 nm using the velocity map ion imaging technique.

Photodissociation dynamics of 1,2-dibromopropane has been investigated at 234 and 265 nm by using the velocity map ion imaging method. At both pump energies, a single Gaussian-shaped speed distribution is observed for the Br*((2)P(1/2)) fragment, whereas at least three velocity components are found to be existent for the Br((2)P(3/2)) product. The secondary C-Br bond cleavage of the bromopropyl radical which is energized from the ultrafast primary C-Br bond rupture should be responsible for the multicomponent translational energy distribution at the low kinetic energy region of Br((2)P(3/2)). The recoil anisotropy parameter (beta) of the fragment from the primary C-Br bond dissociation is measured to be 0.53 (0.49) and 1.26 (1.73) for Br((2)P(3/2)) and Br*((2)P(1/2)), respectively, at 234 (265) nm. The beta value of Br((2)P(3/2)) from the secondary C-Br bond dissociation event at 265 nm is found to be 0.87, reflecting the fact that the corresponding Br((2)P(3/2)) fragment carried the initial vector component of the bromopropyl radical produced from the primary bond dissociation event. Density functional theory has been used to calculate energetics involved both in the primary and in the secondary C-Br bond dissociation dynamics.

The dynamics of Br(2Pj) formation in the photodissociation of vinyl and perfluorovinyl bromides.

The photodissociation dynamics of vinyl bromide and perfluorovinyl bromide have been investigated at 234 nm using a photofragment ion imaging technique coupled with a state-selective [2+1] resonance-enhanced multiphoton ionization scheme. The nascent Br atoms stem from the primary C-Br bond dissociation leading to the formation of C2H3(X) and Br(2Pj;j=1/2,3/2). The obtained translational energy distributions have been well fitted by a single Boltzmann and three Gaussian functions. Boltzmann component has not been observed in the perfluorovinyl bromide. The repulsive 3A'(n,sigma *) state has been considered as the origin of the highest Gaussian components. Middle translational energy components with Gaussian shapes are produced from the 1A"(pi,sigma*) and/or 3A"(pi,sigma*) which are very close in energy. Low-energy Gaussian components are produced via predissociation from the 3A'(pi,pi*) state. The assignments have also been supported by the recoil anisotropy corresponding to the individual components. It is suggested that intersystem crossing from the triplet states to the ground state has been attributed to the Boltzmann component and the fluorination reduces the probability of this electronic relaxation process.