Multimodal protein features play a crucial role in protein function prediction. However, these features encompass a wide range of information, ranging from structural data and sequence features to protein attributes and interaction networks, making it challenging to decipher their complex interconnections. In this work, we propose a multimodal protein function prediction method (DSRPGO) by utilizing dynamic selection and reconstructive pre-training mechanisms. To acquire complex protein information, we introduce reconstructive pre-training to mine more fine-grained information with low semantic levels. Moreover, we put forward the Bidirectional Interaction Module (BInM) to facilitate interactive learning among multimodal features. Additionally, to address the difficulty of hierarchical multi-label classification in this task, a Dynamic Selection Module (DSM) is designed to select the feature representation that is most conducive to current protein function prediction. Our proposed DSRPGO model improves significantly in BPO, MFO, and CCO on human datasets, thereby outperforming other benchmark models.

Identification of intrinsically disordered regions (IDRs) in proteins is essential for understanding fundamental cellular processes. The IDRs can be divided into long disordered regions (LDRs) and short disordered regions (SDRs) according to their lengths. In previous studies, most computational methods ignored the differences between LDRs and SDRs, and therefore failed to capture the different patterns of LDRs and SDRs. In this study, we propose IDP-EDL, an ensemble of three predictors. The component predictors were first built based on pretrained protein language model and applied task-specific fine-tuning for short, long, and generic disordered regions. A meta predictor was then trained to integrate three task-specific predictors into the final predictor. The results of experiments show that task-specific supervised fine-tuning can capture the different features of LDRs and SDRs and IDP-EDL can achieve stable performance on datasets with different ratios of LDRs and SDRs. More importantly, IDP-EDL can reach or even surpass state-of-the-art performance than other existing predictors on independent test sets. IDP-EDL is available at https://github.com/joestarXjx/IDP-EDL.

BACKGROUND: A massive amount of protein sequences have been obtained, but their functions remain challenging to discern. In recent research on protein function prediction, Protein-Protein Interaction (PPI) Networks have played a crucial role. Uncovering potential function relationships between distant proteins within PPI networks is essential for improving the accuracy of protein function prediction. Most current studies attempt to capture these distant relationships by stacking graph network layers, but performance gains diminish as the number of layers increases. RESULTS: To further explore the potential functional relationships between multi-hop proteins in PPI networks, this paper proposes SEGT-GO, a Graph Transformer method based on PPI multi-hop neighborhood Serialization and Explainable artificial intelligence for large-scale multispecies protein function prediction. The multi-hop neighborhood serialization maps multi-hop information in the PPI Network into serialized feature embeddings, enabling the Graph Transformer to learn deeper functional features within the PPI Network. Based on game theory, the SHAP eXplainable Artificial Intelligence (XAI) framework optimizes model input and filters out feature noise, enhancing model performance. CONCLUSIONS: Compared to the advanced network method DeepGraphGO, SEGT-GO achieves more competitive results in standard large-scale datasets and superior results on small ones, validating its ability to extract functional information from deep proteins. Furthermore, SEGT-GO achieves superior results in cross-species learning and prediction of the functions of unseen proteins, further proving the method's strong generalization.

As the most common complication of diabetes, diabetic retinopathy (DR) is one of the main causes of irreversible blindness. Automatic DR grading plays a crucial role in early diagnosis and intervention, reducing the risk of vision loss in people with diabetes. In these years, various deep-learning approaches for DR grading have been proposed. Most previous DR grading models are trained using the dataset of single-field fundus images, but the entire retina cannot be fully visualized in a single field of view. There are also problems of scattered location and great differences in the appearance of lesions in fundus images. To address the limitations caused by incomplete fundus features, and the difficulty in obtaining lesion information. This work introduces a novel multi-view DR grading framework, which solves the problem of incomplete fundus features by jointly learning fundus images from multiple fields of view. Furthermore, the proposed model combines multi-view inputs such as fundus images and lesion snapshots. It utilizes heterogeneous convolution blocks (HCB) and scalable self-attention classes (SSAC), which enhance the ability of the model to obtain lesion information. The experimental results show that our proposed method performs better than the benchmark methods on the large-scale dataset.

Protein function prediction is critical for understanding the cellular physiological and biochemical processes, and it opens up new possibilities for advancements in fields such as disease research and drug discovery. During the past decades, with the exponential growth of protein sequence data, many computational methods for predicting protein function have been proposed. Therefore, a systematic review and comparison of these methods are necessary. In this study, we divide these methods into four different categories, including sequence-based methods, 3D structure-based methods, PPI network-based methods and hybrid information-based methods. Furthermore, their advantages and disadvantages are discussed, and then their performance is comprehensively evaluated and compared. Finally, we discuss the challenges and opportunities present in this field.

Diabetic retinopathy (DR) is the main cause of irreversible blindness for working-age adults. The previous models for DR detection have difficulties in clinical application. The main reason is that most of the previous methods only use single-view data, and the single field of view (FOV) only accounts for about 13% of the FOV of the retina, resulting in the loss of most lesion features. To alleviate this problem, we propose a multi-view model for DR detection, which takes full advantage of multi-view images covering almost all of the retinal field. To be specific, we design a Cross-Interaction Self-Attention based Module (CISAM) that interfuses local features extracted from convolutional blocks with long-range global features learned from transformer blocks. Furthermore, considering the pathological association in different views, we use the feature jigsaw to assemble and learn the features of multiple views. Extensive experiments on the latest public multi-view MFIDDR dataset with 34,452 images demonstrate the superiority of our method, which performs favorably against state-of-the-art models. To the best of our knowledge, this work is the first study on the public large-scale multi-view fundus images dataset for DR detection.

MOTIVATION: Protein function annotation is fundamental to understanding biological mechanisms. The abundant genome-scale protein-protein interaction (PPI) networks, together with other protein biological attributes, provide rich information for annotating protein functions. As PPI networks and biological attributes describe protein functions from different perspectives, it is highly challenging to cross-fuse them for protein function prediction. Recently, several methods combine the PPI networks and protein attributes via the graph neural networks (GNNs). However, GNNs may inherit or even magnify the bias caused by noisy edges in PPI networks. Besides, GNNs with stacking of many layers may cause the over-smoothing problem of node representations. RESULTS: We develop a novel protein function prediction method, CFAGO, to integrate single-species PPI networks and protein biological attributes via a multi-head attention mechanism. CFAGO is first pre-trained with an encoder-decoder architecture to capture the universal protein representation of the two sources. It is then fine-tuned to learn more effective protein representations for protein function prediction. Benchmark experiments on human and mouse datasets show CFAGO outperforms state-of-the-art single-species network-based methods by at least 7.59%, 6.90%, 11.68% in terms of m-AUPR, M-AUPR, and Fmax, respectively, demonstrating cross-fusion by multi-head attention mechanism can greatly improve the protein function prediction. We further evaluate the quality of captured protein representations in terms of Davies Bouldin Score, whose results show that cross-fused protein representations by multi-head attention mechanism are at least 2.7% better than that of original and concatenated representations. We believe CFAGO is an effective tool for protein function prediction. AVAILABILITY AND IMPLEMENTATION: The source code of CFAGO and experiments data are available at: http://bliulab.net/CFAGO/.

Protein remote homology detection is one of the most fundamental research tool for protein structure and function prediction. Most search methods for protein remote homology detection are evaluated based on the Structural Classification of Proteins-extended (SCOPe) benchmark, but the diverse hierarchical structure relationships between the query protein and candidate proteins are ignored by these methods. In order to further improve the predictive performance for protein remote homology detection, a search framework based on the predicted protein hierarchical relationships (PHR-search) is proposed. In the PHR-search framework, the superfamily level prediction information is obtained by extracting the local and global features of the Hidden Markov Model (HMM) profile through a convolution neural network and it is converted to the fold level and class level prediction information according to the hierarchical relationships of SCOPe. Based on these predicted protein hierarchical relationships, filtering strategy and re-ranking strategy are used to construct the two-level search of PHR-search. Experimental results show that the PHR-search framework achieves the state-of-the-art performance by employing five basic search methods, including HHblits, JackHMMER, PSI-BLAST, DELTA-BLAST and PSI-BLASTexB. Furthermore, the web server of PHR-search is established, which can be accessed at http://bliulab.net/PHR-search.

Protein remote homology detection is a fundamental and important task for protein structure and function analysis. Several search methods have been proposed to improve the detection performance of the remote homologues and the accuracy of ranking lists. The position-specific scoring matrix (PSSM) profile and hidden Markov model (HMM) profile can contribute to improving the performance of the state-of-the-art search methods. In this paper, we improved the profile-link (PL) information for constructing PSSM or HMM profiles, and proposed a PL-based search method (PL-search). In PL-search, more robust PLs are constructed through the double-link and iterative extending strategies, and an accurate similarity score of sequence pairs is calculated from the two-level Jaccard distance for remote homologues. We tested our method on two widely used benchmark datasets. Our results show that whether HHblits, JackHMMER or position-specific iterated-BLAST is used, PL-search obviously improves the search performance in terms of ranking quality as well as the number of detected remote homologues. For ease of use of PL-search, both its stand-alone tool and the web server are constructed, which can be accessed at http://bliulab.net/PL-search/.

MOTIVATION: Protein remote homology detection is a challenging task for the studies of protein evolutionary relationships. PSI-BLAST is an important and fundamental search method for detecting homology proteins. Although many improved versions of PSI-BLAST have been proposed, their performance is limited by the search processes of PSI-BLAST. RESULTS: For further improving the performance of PSI-BLAST for protein remote homology detection, a supervised two-layer search framework based on PSI-BLAST (S2L-PSIBLAST) is proposed. S2L-PSIBLAST consists of a two-level search: the first-level search provides high-quality search results by using SMI-BLAST framework and double-link strategy to filter the non-homology protein sequences, the second-level search detects more homology proteins by profile-link similarity, and more accurate ranking lists for those detected protein sequences are obtained by learning to rank strategy. Experimental results on the updated version of Structural Classification of Proteins-extended benchmark dataset show that S2L-PSIBLAST not only obviously improves the performance of PSI-BLAST, but also achieves better performance on two improved versions of PSI-BLAST: DELTA-BLAST and PSI-BLASTexB. AVAILABILITY AND IMPLEMENTATION: http://bliulab.net/S2L-PSIBLAST. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: As one of the most important and widely used mainstream iterative search tool for protein sequence search, an accurate Position-Specific Scoring Matrix (PSSM) is the key of PSI-BLAST. However, PSSMs containing non-homologous information obviously reduce the performance of PSI-BLAST for protein remote homology. RESULTS: To further study this problem, we summarize three types of Incorrectly Selected Homology (ISH) errors in PSSMs. A new search tool Supervised-Manner-based Iterative BLAST (SMI-BLAST) is proposed based on PSI-BLAST for solving these errors. SMI-BLAST obviously outperforms PSI-BLAST on the Structural Classification of Proteins-extended (SCOPe) dataset. Compared with PSI-BLAST on the ISH error subsets of SCOPe dataset, SMI-BLAST detects 1.6-2.87 folds more remote homologous sequences, and outperforms PSI-BLAST by 35.66% in terms of ROC1 scores. Furthermore, this framework is applied to JackHMMER, DELTA-BLAST and PSI-BLASTexB, and their performance is further improved. AVAILABILITY AND IMPLEMENTATION: User-friendly webservers for SMI-BLAST, JackHMMER, DELTA-BLAST and PSI-BLASTexB are established at http://bliulab.net/SMI-BLAST/, by which the users can easily get the results without the need to go through the mathematical details. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.