Interdisciplinary Sciences: Computational Life Sciences最新文献

Computational drug repositioning, through predicting drug-disease associations (DDA), offers significant potential for discovering new drug indications. Current methods incorporate graph neural networks (GNN) on drug-disease heterogeneous networks to predict DDAs, achieving notable performances compared to traditional machine learning and matrix factorization approaches. However, these methods depend heavily on network topology, hampered by incomplete and noisy network data, and overlook the wealth of biomedical knowledge available. Correspondingly, large language models (LLMs) excel in graph search and relational reasoning, which can possibly enhance the integration of comprehensive biomedical knowledge into drug and disease profiles. In this study, we first investigate the contribution of LLM-inferred knowledge representation in drug repositioning and DDA prediction. A zero-shot prompting template was designed for LLM to extract high-quality knowledge descriptions for drug and disease entities, followed by embedding generation from language models to transform the discrete text to continual numerical representation. Then, we proposed LLM-DDA with three different model architectures (LLM-DDA_{Node Feat}, LLM-DDA_{Dual GNN}, LLM-DDA_GNN-AE) to investigate the best fusion mode for LLM-based embeddings. Extensive experiments on four DDA benchmarks show that, LLM-DDA_GNN-AE achieved the optimal performance compared to 11 baselines with the overall relative improvement in AUPR of 23.22%, F1-Score of 17.20%, and precision of 25.35%. Meanwhile, selected case studies of involving Prednisone and Allergic Rhinitis highlighted the model's capability to identify reliable DDAs and knowledge descriptions, supported by existing literature. This study showcases the utility of LLMs in drug repositioning with its generality and applicability in other biomedical relation prediction tasks.

Autism spectrum disorder (ASD) is a serious mental disorder with a complex pathogenesis mechanism and variable presentation among individuals. Although many deep learning algorithms have been used to diagnose ASD, most of them focus on a single modality of data, resulting in limited information extraction and poor stability. In this paper, we propose a bilinear perceptual fusion (BPF) algorithm that leverages data from multiple modalities. In our algorithm, different schemes are used to extract features according to the characteristics of functional and structural data. Through bilinear operations, the associations between the functional and structural features of each region of interest (ROI) are captured. Then the associations are used to integrate the feature representation. Graph convolutional neural networks (GCNs) can effectively utilize topology and node features in brain network analysis. Therefore, we design a deep learning framework called BPF-GCN and conduct experiments on publicly available ASD dataset. The results show that the classification accuracy of BPF-GCN reached 82.35%, surpassing existing methods. This demonstrates the superiority of its classification performance, and the framework can extract ROIs related to ASD. Our work provides a valuable reference for the timely diagnosis and treatment of ASD.

Graphical Abstract

Based on the extracted functional and structural features, we design a generic framework called BPF-GCN. It can not only diagnose ASD, but also identify pathogenic ROIs. BPF-GCN consists of four parts. They are extraction of brain functional features, extraction of brain structural features, feature fusion and classification.

The breakthrough of AlphaFold2 and the publication of AlphaFold DB represent a significant advance in the field of predicting static protein structures. However, AlphaFold2 models tend to represent a single static structure, and multiple-conformation prediction remains a challenge. In this work, we proposed a method named MultiSFold, which uses a distance-based multi-objective evolutionary algorithm to predict multiple conformations. To begin, multiple energy landscapes are constructed using different competing constraints generated by deep learning. Subsequently, an iterative modal exploration and exploitation strategy is designed to sample conformations, incorporating multi-objective optimization, geometric optimization and structural similarity clustering. Finally, the final population is generated using a loop-specific sampling strategy to adjust the spatial orientations. MultiSFold was evaluated against state-of-the-art methods using a benchmark set containing 80 protein targets, each characterized by two representative conformational states. Based on the proposed metric, MultiSFold achieves a remarkable success ratio of 56.25% in predicting multiple conformations, while AlphaFold2 only achieves 10.00%, which may indicate that conformational sampling combined with knowledge gained through deep learning has the potential to generate conformations spanning the range between different conformational states. In addition, MultiSFold was tested on 244 human proteins with low structural accuracy in AlphaFold DB to test whether it could further improve the accuracy of static structures. The experimental results demonstrate the performance of MultiSFold, with a TM-score better than that of AlphaFold2 by 2.97% and RoseTTAFold by 7.72%. The online server is at http://zhanglab-bioinf.com/MultiSFold .

Circular RNAs (circRNAs) are non-coding RNAs generated by reverse splicing. They are involved in biological process and human diseases by interacting with specific RNA-binding proteins (RBPs). Due to traditional biological experiments being costly, computational methods have been proposed to predict the circRNA-RBP interaction. However, these methods have problems of single feature extraction. Therefore, we propose a novel model called circ-FHN, which utilizes only circRNA sequences to predict circRNA-RBP interactions. The circ-FHN approach involves feature coding and a hybrid deep learning model. Feature coding takes into account the physicochemical properties of circRNA sequences and employs four coding methods to extract sequence features. The hybrid deep structure comprises a convolutional neural network (CNN) and a bidirectional gated recurrent unit (BiGRU). The CNN learns high-level abstract features, while the BiGRU captures long-term dependencies in the sequence. To assess the effectiveness of circ-FHN, we compared it to other computational methods on 16 datasets and conducted ablation experiments. Additionally, we conducted motif analysis. The results demonstrate that circ-FHN exhibits exceptional performance and surpasses other methods. circ-FHN is freely available at https://github.com/zhaoqi106/circ-FHN .

The development of therapeutic antibodies is an important aspect of new drug discovery pipelines. The assessment of an antibody's developability-its suitability for large-scale production and therapeutic use-is a particularly important step in this process. Given that experimental assays to assess antibody developability in large scale are expensive and time-consuming, computational methods have been a more efficient alternative. However, the antibody research community faces significant challenges due to the scarcity of readily accessible data on antibody developability, which is essential for training and validating computational models. To address this gap, DOTAD (Database Of Therapeutic Antibody Developability) has been built as the first database dedicated exclusively to the curation of therapeutic antibody developability information. DOTAD aggregates all available therapeutic antibody sequence data along with various developability metrics from the scientific literature, offering researchers a robust platform for data storage, retrieval, exploration, and downloading. In addition to serving as a comprehensive repository, DOTAD enhances its utility by integrating a web-based interface that features state-of-the-art tools for the assessment of antibody developability. This ensures that users not only have access to critical data but also have the convenience of analyzing and interpreting this information. The DOTAD database represents a valuable resource for the scientific community, facilitating the advancement of therapeutic antibody research. It is freely accessible at http://i.uestc.edu.cn/DOTAD/ , providing an open data platform that supports the continuous growth and evolution of computational methods in the field of antibody development.

Protein complex structure prediction is an important problem in computational biology. While significant progress has been made for protein monomers, accurate evaluation of protein complexes remains challenging. Existing assessment methods in CASP, lack dedicated metrics for evaluating complexes. DockQ, a widely used metric, has some limitations. In this study, we propose a novel metric called BDM (Based on Distance difference Matrix) for assessing protein complex prediction structures. Our approach utilizes a distance difference matrix derived from comparing real and predicted protein structures, establishing a linear correlation with Root Mean Square Deviation (RMSD). BDM overcomes limitations associated with receptor-ligand differentiation and eliminates the requirement for structure alignment, making it a more effective and efficient metric. Evaluation of BDM using CASP14 and CASP15 test sets demonstrates superior performance compared to the official CASP scoring. BDM provides accurate and reasonable assessments of predicted protein complexes, wide adoption of BDM has the potential to advance protein complex structure prediction and facilitate related researches across scientific domains. Code is available at http://mialab.ruc.edu.cn/BDMServer/ .

As one of the most important post-translational modifications (PTMs), protein phosphorylation plays a key role in a variety of biological processes. Many studies have shown that protein phosphorylation is associated with various human diseases. Therefore, identifying protein phosphorylation site-disease associations can help to elucidate the pathogenesis of disease and discover new drug targets. Networks of sequence similarity and Gaussian interaction profile kernel similarity were constructed for phosphorylation sites, as well as networks of disease semantic similarity, disease symptom similarity and Gaussian interaction profile kernel similarity were constructed for diseases. To effectively combine different phosphorylation sites and disease similarity information, random walk with restart algorithm was used to obtain the topology information of the network. Then, the diffusion component analysis method was utilized to obtain the comprehensive phosphorylation site similarity and disease similarity. Meanwhile, the reliable negative samples were screened based on the Euclidean distance method. Finally, a convolutional neural network (CNN) model was constructed to identify potential associations between phosphorylation sites and diseases. Based on tenfold cross-validation, the evaluation indicators were obtained including accuracy of 93.48%, specificity of 96.82%, sensitivity of 90.15%, precision of 96.62%, Matthew's correlation coefficient of 0.8719, area under the receiver operating characteristic curve of 0.9786 and area under the precision-recall curve of 0.9836. Additionally, most of the top 20 predicted disease-related phosphorylation sites (19/20 for Alzheimer's disease; 20/16 for neuroblastoma) were verified by literatures and databases. These results show that the proposed method has an outstanding prediction performance and a high practical value.

Mass spectrometry is crucial in proteomics analysis, particularly using Data Independent Acquisition (DIA) for reliable and reproducible mass spectrometry data acquisition, enabling broad mass-to-charge ratio coverage and high throughput. DIA-NN, a prominent deep learning software in DIA proteome analysis, generates peptide results but may include low-confidence peptides. Conventionally, biologists have to manually screen peptide fragment ion chromatogram peaks (XIC) for identifying high-confidence peptides, a time-consuming and subjective process prone to variability. In this study, we introduce SeFilter-DIA, a deep learning algorithm, aiming at automating the identification of high-confidence peptides. Leveraging compressed excitation neural network and residual network models, SeFilter-DIA extracts XIC features and effectively discerns between high and low-confidence peptides. Evaluation of the benchmark datasets demonstrates SeFilter-DIA achieving 99.6% AUC on the test set and 97% for other performance indicators. Furthermore, SeFilter-DIA is applicable for screening peptides with phosphorylation modifications. These results demonstrate the potential of SeFilter-DIA to replace manual screening, providing an efficient and objective approach for high-confidence peptide identification while mitigating associated limitations.

Recognizing drug-target interactions (DTI) stands as a pivotal element in the expansive field of drug discovery. Traditional biological wet experiments, although valuable, are time-consuming and costly as methods. Recently, computational methods grounded in network learning have demonstrated great advantages by effective topological feature extraction and attracted extensive research attention. However, most existing network-based learning methods only consider the low-order binary correlation between individual drug and target, neglecting the potential higher-order correlation information derived from multiple drugs and targets. High-order information, as an essential component, exhibits complementarity with low-order information. Hence, the incorporation of higher-order associations between drugs and targets, while adequately integrating them with the existing lower-order information, could potentially yield substantial breakthroughs in predicting drug-target interactions. We propose a novel dual channels network-based learning model CHL-DTI that converges high-order information from hypergraphs and low-order information from ordinary graph for drug-target interaction prediction. The convergence of high-low order information in CHL-DTI is manifested in two key aspects. First, during the feature extraction stage, the model integrates both high-level semantic information and low-level topological information by combining hypergraphs and ordinary graph. Second, CHL-DTI fully fuse the innovative introduced drug-protein pairs (DPP) hypergraph network structure with ordinary topological network structure information. Extensive experimentation conducted on three public datasets showcases the superior performance of CHL-DTI in DTI prediction tasks when compared to SOTA methods. The source code of CHL-DTI is available at https://github.com/UPCLyy/CHL-DTI .

Sarcomas are malignant tumors from mesenchymal tissue and are characterized by their complexity and diversity. The high recurrence rate making it important to understand the mechanisms behind their recurrence and to develop personalized treatments and drugs. However, previous studies on the association patterns of multi-modal data on sarcoma recurrence have overlooked the fact that genes do not act independently, but rather function within signaling pathways. Therefore, this study collected 290 whole solid images, 869 gene and 1387 pathway data of over 260 sarcoma samples from UCSC and TCGA to identify the association patterns of gene-pathway-cell related to sarcoma recurrences. Meanwhile, considering that most multi-modal data fusion methods based on the joint non-negative matrix factorization (NMF) model led to poor experimental repeatability due to random initialization of factorization parameters, the study proposed the singular value decomposition (SVD)-driven joint NMF model by applying the SVD method to calculate initialized weight and coefficient matrices to achieve the reproducibility of the results. The results of the experimental comparison indicated that the SVD algorithm enhances the performance of the joint NMF algorithm. Furthermore, the representative module indicated a significant relationship between genes in pathways and image features. Multi-level analysis provided valuable insights into the connections between biological processes, cellular features, and sarcoma recurrence. In addition, potential biomarkers were uncovered, while various mechanisms of sarcoma recurrence were identified from an imaging genetic perspective. Overall, the SVD-NMF model affords a novel perspective on combining multi-omics data to explore the association related to sarcoma recurrence.