The precise 3-D ionospheric electron density (IED) modeling is in general difficult due to costly and insufficient observations in the ionosphere. The Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) mission has provided a significant opportunity for estimating the IED profile during the past 15 years. In this article, a precise empirical 3-D global electron density model (GEDM) from COSMIC GNSS radio occultation (RO) data is developed based on Fourier expansion and principal component analysis (PCA) and evaluated with independent datasets. Each profile is described by five essential parameters: the mean scale height Hm, bottom and topside slopes of the scale height (a1 and a2), peak density of the F2 layer (NmF2), and height of the F2 layer peak density (hmF2). The GEDM model can provide IED at any local time (LT) (0–24 h), F10.7 (70–150 sfu), Kp (0–9), latitude (90°S–90°N), longitude (180°W–180°E), and altitude, especially in the F region. The model is validated by the incoherent scatter radar (ISR), International Reference Ionosphere (IRI)-2020, K-Band Ranging system (KBR) of the Gravity Recovery and Climate Experiment (GRACE), and the Global Ionospheric Map (GIM) TEC in 2010 and 2014. The results show that the GEDM model has preferable reliability and consistency with respect to the independent observations and the IRI-2020 model. The statistical error of the GEDM model is lower than that of the IRI-2020, particularly in the mid-latitude region. Furthermore, the GEDM model is capable of precisely capturing typical ionospheric features, such as the equatorial ionization anomaly (EIA), winter anomaly, and the annual anomaly.

Large scale RNA-protein binding data made computational identification of RNA-protein interactions possible, which provides great convenience for revealing the interplay between non-coding RNAs and RNA-binding proteins (RBPs). Various machine learning methods have been applied to the prediction of RNA-protein interactions (RPIs). However, most of the RPI predictors were designed for linear RNAs, while methods for circular RNAs (circRNAs) are few. Besides, most of the predictors use only RNA sequence or only protein sequences as input. Automatic feature learning from RNA and RBP sequences using deep neural networks has great potential to improve RPI predictors. In this paper, we propose a new method, PSi-bind, to predict the binding relationship between RNAs and RBPs, which is featured by a pseudo-Siamese neural network. We feed both RNA and RBP sequence embedding features to the network and perform an end-to-end learning to discover the latent association between RNAs and RBPs. Especially, we train our model on a large-scale circRNA dataset. The experimental results show that the model achieves very high accuracy on independent test set, and outperforms the state-of-the-art RPI predictors by a large margin.

Monitoring and understanding geophysical processes in the thermosphere is primordial for space physics, low Earth orbiters, and ground-based technologies. In the last half century, thermospheric variations, anomalies, and climatology have been investigated and reported, but were limited due to lack of observations and large uncertainty in the models. Today, Global Navigation Satellite Systems (GNSS) and accelerometers on small satellites can sense neutraldensity and wind variations with unprecedented accuracy, which contribute to understand thermospheric variations and improve the current empirical and physical models. In this paper, an overview of past and present developments and efforts in sensing and modeling thermospheric density and wind variations is presented, as well as the future challenges and perspectives for GNSS and accelerometers on small satellites.