Genomic prediction (GP) has made significant progress in the field of breeding. Traditional linear models perform well in handling simple traits but have limitations in extracting nonlinear features for complex traits. The introduction of deep learning (DL) techniques has provided a new approach to GP, especially suited for high-dimensional data processing and complex trait prediction. However, traditional DL models require manual design of the network architecture, which necessitates continuous experimentation and modification. In this paper, we propose a new framework, MDNN, that utilizes the memetic algorithm for neural architecture search and automatically optimizes the network architecture. Compared with the DNNGP, MDNN achieved a 36.49% improvement in the average Pearson correlation coefficient on the wheat599 dataset and a 12.28% improvement on the wheat2000 dataset.

Predicting long non-coding RNA (lncRNA)-protein interactions is essential for understanding biological processes and discovering new therapeutic targets. In this study, we propose a novel model based on inter-view contrastive learning and miRNA fusion for lncRNA-protein interaction (LPI) prediction, called ICMF-LPI, which utilizes a heterogeneous information network to enhance LPI prediction. The model integrates miRNA as a mediator, constructing an lncRNA-miRNA-protein network, and employs metapath to extract diverse relationships from heterogeneous graphs. By fusing miRNA-related information and leveraging contrastive learning across inter-views, ICMF-LPI effectively captures potential interactions. Experimental results, including five-fold cross-validation, demonstrate the model's superior performance compared to several state-of-the-art methods, with significant improvements in the area under the receiver operating characteristic curve and the area under the precision-recall curve metrics. Notably, even when direct LPI connections are excluded, ICMF-LPI still achieves competitive predictive accuracy, performing comparably or better than some existing models. This demonstrates that the proposed model is effective in scenarios where direct interaction data are unavailable. This approach offers a promising direction for developing predictive models in bioinformatics, particularly in challenging conditions.

Accurately predicting the functions of multi-functional therapeutic peptides is crucial for the development of related drugs. However, existing peptide function prediction methods largely rely on either a single type of feature or a single model architecture, limiting prediction accuracy and applicability. Additionally, training better-performing models on datasets with class imbalance issues remains a significant challenge. In this study, we propose the multi-functional therapeutic peptide of multi-feature fusion prediction (MFTP_MFFP) model, a novel method for predicting the functionality of multi-functional therapeutic peptides. This approach uses various encoding techniques to process peptide sequence data, generating multiple features that help the model learn hidden information within the sequences. To maximize the effectiveness of these features, we propose a gated feature fusion module that efficiently integrates them. The module assigns learnable gating weights to each feature, optimizing integration and enhancing fusion efficiency. The fused features are then passed into a neural network model for feature extraction. Additionally, we propose a marginal focal dice loss function (MFDL) to address the class imbalance and improve the model's prediction performance. Experimental results show that the MFTP_MFFP model outperforms existing models in all evaluation metrics, demonstrating its robustness and effectiveness in multi-functional therapeutic peptide prediction tasks.

This paper proposes a deep learning (DL)-based time series (TS) anomaly detection method (DLTS) for the low earth orbit (LEO) satellite network slicing scenario, aiming to address the virtual link anomalies induced by software and hardware abnormalities. Initially, the time series anomalous variations of each resource utilization in satellite network slicing are categorized into three types based on the utilization of computing, storage, and network resources of virtual nodes. Thereafter, the anomaly detection problem is formulated as a classification problem, and the time series are transformed into images using the Gramian Angular Field (GAF) for model input. Lastly, we propose a design principle for a time-constrained deep neural network architecture to mitigate training time, and design a DL model architecture to classify the TS transformation images of resource utilization for each virtual node. This aligns with the objective of the satellite network slicing scenario. Additionally, a new evaluation metric is introduced. Experimental results underscore the shorter training time of the proposed model, and affirm its efficacy, demonstrated through accuracy, F1 score, and the newly proposed evaluation metric.