An Augmented Sample Selection Framework for Prediction of Anticancer Peptides.

Anticancer peptides (ACPs) have promising prospects for cancer treatment. Traditional ACP identification experiments have the limitations of low efficiency and high cost. In recent years, data-driven deep learning techniques have shown significant potential for ACP prediction. However, data-driven prediction models rely heavily on extensive training data. Furthermore, the current publicly accessible ACP dataset is limited in size, leading to inadequate model generalization. While data augmentation effectively expands dataset size, existing techniques for augmenting ACP data often generate noisy samples, adversely affecting prediction performance. Therefore, this paper proposes a novel augmented sample selection framework for the prediction of anticancer peptides (ACPs-ASSF). First, the prediction model is trained using raw data. Then, the augmented samples generated using the data augmentation technique are fed into the trained model to compute pseudo-labels and estimate the uncertainty of the model prediction. Finally, samples with low uncertainty, high confidence, and pseudo-labels consistent with the original labels are selected and incorporated into the training set to retrain the model. The evaluation results for the ACP240 and ACP740 datasets show that ACPs-ASSF achieved accuracy improvements of up to 5.41% and 5.68%, respectively, compared to the traditional data augmentation method.

In situ encapsulation of iron oxide nanoparticles into nitrogen-doped carbon nanotubes as anodic electrode materials of lithium ion batteries.

Fe-based oxides are considered as promising anode materials for lithium-ion batteries (LIBs) due to their high theoretical capacities, low cost, natural abundance and environmental friendliness. However, their severe volume expansion upon cycling and poor conductivity limit their cycling stability and rate capability. To address this issue, a hybrid of Fe2O3 nanoparticles encapsulated at the endpoints of nitrogen-doped CNTs (Fe2O3@NCNTs) is designed and prepared using a metal-catalyzed graphitization-nitridization driven tip-growth process and subsequent oxidation in air. When evaluated as an anode material for LIBs, this Fe2O3@NCNT hybrid exhibits a high capacity of 1145 mA h g-1 at 100 mA g-1, excellent rate capability of 907 mA h g-1 at 5.0 A g-1 and remarkable cycling stability of 856 mA h g-1 after 800 cycles at 1 A g-1, which are much superior to those of the Fe2O3/carbon black (CB) control material. The outstanding electrochemical performance benefits from the unique nanoarchitecture of Fe2O3@NCNTs, which provides a porous conductive matrix for effective electron-ion transport, and provides space confining carbon nanocaps as well as stress buffer nanocavities for robust structural stability during the lithiation/delithiation process. The results may pave the way for the rational structural design of high-performance metal oxide-based anode materials for next-generation LIBs.

Multi-Stream Convolution-Recurrent Neural Networks Based on Attention Mechanism Fusion for Speech Emotion Recognition.

The quality of feature extraction plays a significant role in the performance of speech emotion recognition. In order to extract discriminative, affect-salient features from speech signals and then improve the performance of speech emotion recognition, in this paper, a multi-stream convolution-recurrent neural network based on attention mechanism (MSCRNN-A) is proposed. Firstly, a multi-stream sub-branches full convolution network (MSFCN) based on AlexNet is presented to limit the loss of emotional information. In MSFCN, sub-branches are added behind each pooling layer to retain the features of different resolutions, different features from which are fused by adding. Secondly, the MSFCN and Bi-LSTM network are combined to form a hybrid network to extract speech emotion features for the purpose of supplying the temporal structure information of emotional features. Finally, a feature fusion model based on a multi-head attention mechanism is developed to achieve the best fusion features. The proposed method uses an attention mechanism to calculate the contribution degree of different network features, and thereafter realizes the adaptive fusion of different network features by weighting different network features. Aiming to restrain the gradient divergence of the network, different network features and fusion features are connected through shortcut connection to obtain fusion features for recognition. The experimental results on three conventional SER corpora, CASIA, EMODB, and SAVEE, show that our proposed method significantly improves the network recognition performance, with a recognition rate superior to most of the existing state-of-the-art methods.

Strong Generalized Speech Emotion Recognition Based on Effective Data Augmentation.

The absence of labeled samples limits the development of speech emotion recognition (SER). Data augmentation is an effective way to address sample sparsity. However, there is a lack of research on data augmentation algorithms in the field of SER. In this paper, the effectiveness of classical acoustic data augmentation methods in SER is analyzed, based on which a strong generalized speech emotion recognition model based on effective data augmentation is proposed. The model uses a multi-channel feature extractor consisting of multiple sub-networks to extract emotional representations. Different kinds of augmented data that can effectively improve SER performance are fed into the sub-networks, and the emotional representations are obtained by the weighted fusion of the output feature maps of each sub-network. And in order to make the model robust to unseen speakers, we employ adversarial training to generalize emotion representations. A discriminator is used to estimate the Wasserstein distance between the feature distributions of different speakers and to force the feature extractor to learn the speaker-invariant emotional representations by adversarial training. The simulation experimental results on the IEMOCAP corpus show that the performance of the proposed method is 2-9% ahead of the related SER algorithm, which proves the effectiveness of the proposed method.

Spatiotemporal characterization of microdamage accumulation in rat ulnae in response to uniaxial compressive fatigue loading.

Repetitive fatigue loading can induce microdamage accumulation in bone matrix, which results in impaired mechanical properties and increased fracture susceptibility. However, the spatial distribution and time-variant process of microdamage accumulation in fatigue-loaded skeleton, especially for linear microcracks which are known to initiate bone remodeling, remain not fully understood. In this study, the time-varying process of the morphology and distribution of microcracks in rat ulnae subjected to uniaxial compressive fatigue loading was investigated. Right forelimbs of thirty four-month-old male Sprague-Dawley rats were subjected to one bout of cyclic ramp loading with 0.67 Hz at a normalized peak force of 0.055 N/g body weight for 6000 cycles, and the contralateral left ulnae were not loaded as the control samples. Ten rats were randomly euthanized on Days 3, 5, and 7 post fatigue loading. Our findings via two-dimensional histomorphometric measurements based on basic fuchsin staining and three-dimensional quantifications using contrast-enhanced micro-computed tomography (MicroCT) with precipitated BaSO4 staining demonstrated that the accumulation of linear microcracks (increase in the amount of linear microcracks) on Day 5 was significantly higher than that on Day 3 and Day 7 post fatigue loading. Our histological and histomorphometric results revealed that linear microcrack density (Cr.Dn) in the tensile cortex at Days 3, 5 and 7 post fatigue loading was significantly higher than that in the compressive side, whereas linear microcrack length (Cr.Le) in the tensile cortex at Day 3 was significantly lower than that in the compressive cortex. Our findings revealed that microcrack accumulation exhibited a non-linear time-varying process at 3, 5 and 7 days post axial compressive fatigue loading (with observable peak Cr.Dn at Day 5). Our findings also revealed distinct distribution of microcrack density and morphology in rat ulnae with tensile and compressive strains, as characterized by more microcracks accumulated in tensile cortices, and longer cracks shown in compressive cortices.

Pulsed electromagnetic fields (PEMF) attenuate changes in vertebral bone mass, architecture and strength in ovariectomized mice.

Pulsed electromagnetic fields (PEMF) has been investigated as a noninvasive alternative method to prevent bone loss for postmenopausal osteoporosis (OP), and the bone tissue involved in these studies are usually long bones such as femur and tibia in OP patients or rat models. However, few studies have investigated the effects of PEMF on the vertebral bone in mice with OP. This study aimed to investigate whether PEMF preserve lumbar vertebral bone mass, microarchitecture and strength in ovariectomized (OVX) mouse model of OP and its associated mechanisms. Thirty 3-month-old female BALB/c mice were randomly divided into three groups (n=10): sham-operated control (Sham), ovariectomy (OVX), and ovariectomy with PEMF treatment (OVX+PEMF). The OVX+PEMF group was exposed to 15Hz, 1.6 mT PEMF for 8h/day, 7days/week. After 8weeks, the mice were sacrificed. The OVX+PEMF group showed lower body weight gain of mice induced by estrogen deficiency compared with OVX group. Biochemical analysis of serum demonstrated that serum bone formation markers including bone specific alkaline phosphatase (BALP), serum osteocalcin (OCN), osteoprotegerin (OPG) and N-terminal propeptide of type I procollagen (P1NP) were markedly higher in OVX+PEMF group compared with OVX group. Besides, serum bone resorption markers including tartrate-resistant acid phosphatase 5b (TRAP-5b) and C-terminal crosslinked telopeptides of type I collagen (CTX-I) were markedly lower in OVX+PEMF group compared with OVX group. Biomechanical test observed that OVX+PEMF group showed higher compressive maximum load and stiffness of the lumbar vertebrae compared with OVX group. Micro-computed tomography (µCT) and histological analysis of lumbar vertebrae revealed that PEMF partially prevented OVX-induced decrease of trabecular bone mass and deterioration of trabecular bone microarchitecture in lumbar vertebrae. Real-time PCR showed that the canonical Wnt signaling pathway of the lumbar vertebrae, including Wnt3a, LRP5 and ß-catenin were markedly up-regulated in OVX+PEMF group compared with OVX group. Moreover, the mRNA expressions of RANKL and OPG were markedly up-regulated in OVX+PEMF group compared with OVX group, whereas no statistical difference in RANKL/OPG mRNA ratio was found between OVX+PEMF group and OVX group. Besides, our study also found that the RANK mRNA expression was down-regulated in OVX+PEMF group compared with OVX group. Taken together, we reported that long-term stimulation with PEMF treatment was able to alleviate lumbar vertebral OP in postmenopausal mice through a combination of increased bone formation and suppressed bone resorption related to regulating the skeletal gene expressions of Wnt3a/LRP5/ß-catenin and OPG/RANKL/RANK signaling pathways.

Effects of four kinds of electromagnetic fields (EMF) with different frequency spectrum bands on ovariectomized osteoporosis in mice.

Electromagnetic fields (EMF) was considered as a non-invasive modality for treatment of osteoporosis while the effects were diverse with EMF parameters in time domain. In present study, we extended analysis of EMF characteristics from time domain to frequency domain, aiming to investigate effects of four kinds of EMF (LP (1-100 Hz), BP (100-3,000 Hz), HP (3,000-50,000 Hz) and AP (1-50,000 Hz)) on ovariectomized (OVX) osteoporosis (OP) in mice. Forty-eight 3-month-old female BALB/c mice were equally assigned to Sham, OVX, OVX + LP, OVX + BP, OVX + HP and OVX + AP groups (n = 8). After 8-week exposure (3 h/day), LP and BP significantly increased serum bone formation markers and osteogenesis-related gene expressions compared with OVX. Bedsides, LP and BP also slightly increased bone resorption activity compared with OVX, evidenced by increased RANKL/OPG ratio. HP sharply decreased serum bone formation and resporption markers and osteogenesis and osteoclastogenesis related gene expressions compared with OVX. AP had accumulative effects of LP, BP and HP, which significantly increased bone formation and decreased bone resporption activity compared with OVX. As a result, LP, BP and HP exposure did not later deterioration of bone mass, microarchitecture and mechanical strength in OVX mice with OP. However, AP stimulation attenuated OVX-induced bone loss.