ULDNA: integrating unsupervised multi-source language models with LSTM-attention network for high-accuracy protein-DNA binding site prediction.

Efficient and accurate recognition of protein-DNA interactions is vital for understanding the molecular mechanisms of related biological processes and further guiding drug discovery. Although the current experimental protocols are the most precise way to determine protein-DNA binding sites, they tend to be labor-intensive and time-consuming. There is an immediate need to design efficient computational approaches for predicting DNA-binding sites. Here, we proposed ULDNA, a new deep-learning model, to deduce DNA-binding sites from protein sequences. This model leverages an LSTM-attention architecture, embedded with three unsupervised language models that are pre-trained on large-scale sequences from multiple database sources. To prove its effectiveness, ULDNA was tested on 229 protein chains with experimental annotation of DNA-binding sites. Results from computational experiments revealed that ULDNA significantly improves the accuracy of DNA-binding site prediction in comparison with 17 state-of-the-art methods. In-depth data analyses showed that the major strength of ULDNA stems from employing three transformer language models. Specifically, these language models capture complementary feature embeddings with evolution diversity, in which the complex DNA-binding patterns are buried. Meanwhile, the specially crafted LSTM-attention network effectively decodes evolution diversity-based embeddings as DNA-binding results at the residue level. Our findings demonstrated a new pipeline for predicting DNA-binding sites on a large scale with high accuracy from protein sequence alone.

GMFGRN: a matrix factorization and graph neural network approach for gene regulatory network inference.

The recent advances of single-cell RNA sequencing (scRNA-seq) have enabled reliable profiling of gene expression at the single-cell level, providing opportunities for accurate inference of gene regulatory networks (GRNs) on scRNA-seq data. Most methods for inferring GRNs suffer from the inability to eliminate transitive interactions or necessitate expensive computational resources. To address these, we present a novel method, termed GMFGRN, for accurate graph neural network (GNN)-based GRN inference from scRNA-seq data. GMFGRN employs GNN for matrix factorization and learns representative embeddings for genes. For transcription factor-gene pairs, it utilizes the learned embeddings to determine whether they interact with each other. The extensive suite of benchmarking experiments encompassing eight static scRNA-seq datasets alongside several state-of-the-art methods demonstrated mean improvements of 1.9 and 2.5% over the runner-up in area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC). In addition, across four time-series datasets, maximum enhancements of 2.4 and 1.3% in AUROC and AUPRC were observed in comparison to the runner-up. Moreover, GMFGRN requires significantly less training time and memory consumption, with time and memory consumed <10% compared to the second-best method. These findings underscore the substantial potential of GMFGRN in the inference of GRNs. It is publicly available at https://github.com/Lishuoyy/GMFGRN.

MLSNet: a deep learning model for predicting transcription factor binding sites.

Accurate prediction of transcription factor binding sites (TFBSs) is essential for understanding gene regulation mechanisms and the etiology of diseases. Despite numerous advances in deep learning for predicting TFBSs, their performance can still be enhanced. In this study, we propose MLSNet, a novel deep learning architecture designed specifically to predict TFBSs. MLSNet innovatively integrates multisize convolutional fusion with long short-term memory (LSTM) networks to effectively capture DNA-sparse higher-order sequence features. Further, MLSNet incorporates super token attention and Bi-LSTM to systematically extract and integrate higher-order DNA shape features. Experimental results on 165 ChIP-seq (chromatin immunoprecipitation followed by sequencing) datasets indicate that MLSNet consistently outperforms several state-of-the-art algorithms in the prediction of TFBSs. Specifically, MLSNet reports average metrics: 0.8306 for ACC, 0.8992 for AUROC, and 0.9035 for AUPRC, surpassing the second-best methods by 1.82%, 1.68%, and 1.54%, respectively. This research delineates the effectiveness of combining multi-size convolutional layers with LSTM and DNA shape-based features in enhancing predictive accuracy. Moreover, this study comprehensively assesses the variability in model performance across different cell lines and transcription factors. The source code of MLSNet is available at https://github.com/minghaidea/MLSNet.

TripletCell: a deep metric learning framework for accurate annotation of cell types at the single-cell level.

Single-cell RNA sequencing (scRNA-seq) has significantly accelerated the experimental characterization of distinct cell lineages and types in complex tissues and organisms. Cell-type annotation is of great importance in most of the scRNA-seq analysis pipelines. However, manual cell-type annotation heavily relies on the quality of scRNA-seq data and marker genes, and therefore can be laborious and time-consuming. Furthermore, the heterogeneity of scRNA-seq datasets poses another challenge for accurate cell-type annotation, such as the batch effect induced by different scRNA-seq protocols and samples. To overcome these limitations, here we propose a novel pipeline, termed TripletCell, for cross-species, cross-protocol and cross-sample cell-type annotation. We developed a cell embedding and dimension-reduction module for the feature extraction (FE) in TripletCell, namely TripletCell-FE, to leverage the deep metric learning-based algorithm for the relationships between the reference gene expression matrix and the query cells. Our experimental studies on 21 datasets (covering nine scRNA-seq protocols, two species and three tissues) demonstrate that TripletCell outperformed state-of-the-art approaches for cell-type annotation. More importantly, regardless of protocols or species, TripletCell can deliver outstanding and robust performance in annotating different types of cells. TripletCell is freely available at https://github.com/liuyan3056/TripletCell. We believe that TripletCell is a reliable computational tool for accurately annotating various cell types using scRNA-seq data and will be instrumental in assisting the generation of novel biological hypotheses in cell biology.

VPatho: a deep learning-based two-stage approach for accurate prediction of gain-of-function and loss-of-function variants.

Determining the pathogenicity and functional impact (i.e. gain-of-function; GOF or loss-of-function; LOF) of a variant is vital for unraveling the genetic level mechanisms of human diseases. To provide a 'one-stop' framework for the accurate identification of pathogenicity and functional impact of variants, we developed a two-stage deep-learning-based computational solution, termed VPatho, which was trained using a total of 9619 pathogenic GOF/LOF and 138 026 neutral variants curated from various databases. A total number of 138 variant-level, 262 protein-level and 103 genome-level features were extracted for constructing the models of VPatho. The development of VPatho consists of two stages: (i) a random under-sampling multi-scale residual neural network (ResNet) with a newly defined weighted-loss function (RUS-Wg-MSResNet) was proposed to predict variants' pathogenicity on the gnomAD_NV + GOF/LOF dataset; and (ii) an XGBOD model was constructed to predict the functional impact of the given variants. Benchmarking experiments demonstrated that RUS-Wg-MSResNet achieved the highest prediction performance with the weights calculated based on the ratios of neutral versus pathogenic variants. Independent tests showed that both RUS-Wg-MSResNet and XGBOD achieved outstanding performance. Moreover, assessed using variants from the CAGI6 competition, RUS-Wg-MSResNet achieved superior performance compared to state-of-the-art predictors. The fine-trained XGBOD models were further used to blind test the whole LOF data downloaded from gnomAD and accordingly, we identified 31 nonLOF variants that were previously labeled as LOF/uncertain variants. As an implementation of the developed approach, a webserver of VPatho is made publicly available at http://csbio.njust.edu.cn/bioinf/vpatho/ to facilitate community-wide efforts for profiling and prioritizing the query variants with respect to their pathogenicity and functional impact.

MINDG: a drug-target interaction prediction method based on an integrated learning algorithm.

MOTIVATION: Drug-target interaction (DTI) prediction refers to the prediction of whether a given drug molecule will bind to a specific target and thus exert a targeted therapeutic effect. Although intelligent computational approaches for drug target prediction have received much attention and made many advances, they are still a challenging task that requires further research. The main challenges are manifested as follows: (i) most graph neural network-based methods only consider the information of the first-order neighboring nodes (drug and target) in the graph, without learning deeper and richer structural features from the higher-order neighboring nodes. (ii) Existing methods do not consider both the sequence and structural features of drugs and targets, and each method is independent of each other, and cannot combine the advantages of sequence and structural features to improve the interactive learning effect. RESULTS: To address the above challenges, a Multi-view Integrated learning Network that integrates Deep learning and Graph Learning (MINDG) is proposed in this study, which consists of the following parts: (i) a mixed deep network is used to extract sequence features of drugs and targets, (ii) a higher-order graph attention convolutional network is proposed to better extract and capture structural features, and (iii) a multi-view adaptive integrated decision module is used to improve and complement the initial prediction results of the above two networks to enhance the prediction performance. We evaluate MINDG on two dataset and show it improved DTI prediction performance compared to state-of-the-art baselines. AVAILABILITY AND IMPLEMENTATION: https://github.com/jnuaipr/MINDG.

Spin Hall-induced bilinear magnetoelectric resistance.

Magnetoresistance is a fundamental transport phenomenon that is essential for reading the magnetic states for various information storage, innovative computing and sensor devices. Recent studies have expanded the scope of magnetoresistances to the nonlinear regime, such as a bilinear magnetoelectric resistance (BMER), which is proportional to both electric field and magnetic field. Here we demonstrate that the BMER is a general phenomenon that arises even in three-dimensional systems without explicit momentum-space spin textures. Our theory suggests that the spin Hall effect enables the BMER provided that the magnitudes of spin accumulation at the top and bottom interfaces are not identical. The sign of the BMER follows the sign of the spin Hall effect of heavy metals, thereby evidencing that the BMER originates from the bulk spin Hall effect. Our observation suggests that the BMER serves as a general nonlinear transport characteristic in three-dimensional systems, especially playing a crucial role in antiferromagnetic spintronics.

Leucine-Rich Repeat Containing 15-Mediated Cell Adhesion Is Essential for Integrin Signaling in TGF-ß1-Induced PDL Fibroblastic Differentiation.

Human periodontal ligament cells (hPDLCs) cultured from periodontal ligament (PDL) tissue contain postnatal stem cells that can be differentiated into PDL fibroblasts. We obtained PDL fibroblasts from hPDLCs by treatment with low concentrations of TGF-ß1. Since the extracellular matrix and cell surface molecules play an important role in differentiation, we had previously developed a series of monoclonal antibodies against PDL fibroblast-specific cell surface molecules. One of these, the anti-PDL51 antibody, recognized a protein that was significantly upregulated in TGF-ß1-induced PDL fibroblasts and highly accumulated in the PDL region of the tooth root. Mass spectrometry revealed that the antigen recognized by the anti-PDL51 antibody was leucine-rich repeat containing 15 (LRRC15), and this antibody specifically recognized the extracellular glycosylated moiety of LRRC15. Experiments presented here show that as fibroblastic differentiation progresses, increased amounts of LRRC15 localized at the cell surface and membrane. Inhibition of LRRC15 by siRNA-mediated depletion and by antibody blocking resulted in downregulation of the representative PDL fibroblastic markers. Moreover, following LRRC15 inhibition, the directed and elongated cell phenotypes disappeared, and the long processes of the end of the cell body were no longer found. Through a specific interaction between integrin ß1 and LRRC15, the focal adhesion kinase signaling pathway was activated in PDL fibroblasts. Furthermore, it was shown that increased LRRC15 was important for the activation of the integrin-mediated cell adhesion signal pathway for regulation of cellular functions, including fibroblastic differentiation, proliferation, and cell migration arising from the expression of PDL-related genes in TGF-ß1-induced PDL fibroblastic differentiation.

Epigenomic Profiling of Human CD4+ T Cells Supports a Linear Differentiation Model and Highlights Molecular Regulators of Memory Development.

The impact of epigenetics on the differentiation of memory T (Tmem) cells is poorly defined. We generated deep epigenomes comprising genome-wide profiles of DNA methylation, histone modifications, DNA accessibility, and coding and non-coding RNA expression in naive, central-, effector-, and terminally differentiated CD45RA+ CD4+ Tmem cells from blood and CD69+ Tmem cells from bone marrow (BM-Tmem). We observed a progressive and proliferation-associated global loss of DNA methylation in heterochromatic parts of the genome during Tmem cell differentiation. Furthermore, distinct gradually changing signatures in the epigenome and the transcriptome supported a linear model of memory development in circulating T cells, while tissue-resident BM-Tmem branched off with a unique epigenetic profile. Integrative analyses identified candidate master regulators of Tmem cell differentiation, including the transcription factor FOXP1. This study highlights the importance of epigenomic changes for Tmem cell biology and demonstrates the value of epigenetic data for the identification of lineage regulators.

Defining the activity of pro-reparative extracellular vesicles in wound healing based on miRNA payloads and cell type-specific lineage mapping.

Small extracellular vesicles (EVs) are released by cells and deliver biologically active payloads to coordinate the response of multiple cell types in cutaneous wound healing. Here we used a cutaneous injury model as a donor of pro-reparative EVs to treat recipient diabetic obese mice, a model of impaired wound healing. We established a functional screen for microRNAs (miRNAs) that increased the pro-reparative activity of EVs and identified a down-regulation of miR-425-5p in EVs in vivo and in vitro associated with the regulation of adiponectin. We tested a cell type-specific reporter of a tetraspanin CD9 fusion with GFP to lineage map the release of EVs from macrophages in the wound bed, based on the expression of miR-425-5p in macrophage-derived EVs and the abundance of macrophages in EV donor sites. Analysis of different promoters demonstrated that EV release under the control of a macrophage-specific promoter was most abundant and that these EVs were internalized by dermal fibroblasts. These findings suggested that pro-reparative EVs deliver miRNAs, such as miR-425-5p, that stimulate the expression of adiponectin that has insulin-sensitizing properties. We propose that EVs promote intercellular signaling between cell layers in the skin to resolve inflammation, induce proliferation of basal keratinocytes, and accelerate wound closure.

Sex pheromone communication in an insect parasitoid, Campoletis chlorideae Uchida.

Sex pheromones are pivotal for insect reproduction. However, the mechanism of sex pheromone communication remains enigmatic in hymenopteran parasitoids. Here we have identified the sex pheromone and elucidated the olfactory basis of sex pheromone communication in Campoletis chlorideae (Ichneumonidae), a solitary larval endoparasitoid of over 30 lepidopteran pests. Using coupled gas chromatography-electroantennogram detection, we identified two female-derived pheromone components, tetradecanal (14:Ald) and 2-heptadecanone (2-Hep) (1:4.6), eliciting strong antennal responses from males but weak responses from females. We observed that males but not females were attracted to both single components and the blend. The hexane-washed female cadavers failed to arouse males, and replenishing 14:Ald and 2-Hep could partially restore the sexual attraction of males. We further expressed six C. chlorideae male-biased odorant receptors in Drosophila T1 neurons and found that CchlOR18 and CchlOR47 were selectively tuned to 14:Ald and 2-Hep, respectively. To verify the biological significance of this data, we knocked down CchlOR18 and CchlOR47 individually or together in vivo and show that the attraction of C. chlorideae to their respective ligands was abolished. Moreover, the parasitoids defective in either of the receptors were less likely to court and copulate. Finally, we show that the sex pheromone and (Z)-jasmone, a potent female attractant, can synergistically affect behaviors of virgin males and virgin females and ultimately increase the parasitic efficiency of C. chlorideae. Our study provides new insights into the molecular mechanism of sex pheromone communication in C. chlorideae that may permit manipulation of parasitoid behavior for pest control.

Phase 1/2 trials of human bone marrow-derived clonal mesenchymal stem cells for treatment of adults with moderate to severe atopic dermatitis.

BACKGROUND: Mesenchymal stem cells (MSCs) play important roles in therapeutic applications by regulating immune responses. OBJECTIVE: We investigated the safety and efficacy of allogenic human bone marrow-derived clonal MSCs (hcMSCs) in subjects with moderate to severe atopic dermatitis (AD). METHODS: The study included a phase 1 open-label trial followed by a phase 2 randomized, double-blind, placebo-controlled trial that involved 72 subjects with moderate to severe AD. RESULTS: In phase 1, intravenous administration of hcMSCs at 2 doses (1 × 106 and 5 × 105 cells/kg) was safe and well tolerated in 20 subjects. Because there was no difference between the 2 dosage groups (P = .9), it was decided to administer low-dose hcMSCs only for phase 2. In phase 2, subjects receiving 3 weekly intravenous infusions of hcMSCs at 5 × 105 cells/kg showed a higher proportion of an Eczema Area and Severity Index (EASI)-50 response at week 12 compared to the placebo group (P = .038). The differences between groups in the Dermatology Life Quality Index and pruritus numeric rating scale scores were not statistically significant. Most adverse events were mild or moderate and resolved by the end of the study period. CONCLUSIONS: The hcMSC treatment resulted in a significantly higher rate of EASI-50 at 12 weeks compared to the control group in subjects with moderate to severe AD. The safety profile of hcMSC treatment was acceptable. Further larger-scale studies are necessary to confirm these preliminary findings.

Glioma stem cell-derived exosomes induce the transformation of astrocytes via the miR-3065-5p/DLG2 signaling axis.

Tumor-associated astrocytes (TAAs) in the glioblastoma microenvironment play an important role in tumor development and malignant progression initiated by glioma stem cells (GSCs). In the current study, normal human astrocytes (NHAs) were cultured and continuously treated with GSC-derived exosomes (GSC-EXOs) induction to explore the mechanism by which GSCs affect astrocyte remodeling. This study revealed that GSC-EXOs can induce the transformation of NHAs into TAAs, with relatively swollen cell bodies and multiple extended processes. In addition, high proliferation, elevated resistance to temozolomide (TMZ), and increased expression of TAA-related markers (TGF-ß, CD44, and tenascin-C) were observed in the TAAs. Furthermore, GSC-derived exosomal miR-3065-5p could be delivered to NHAs, and miR-3065-5p levels increased significantly in TAAs, as verified by miRNA expression profile sequencing and Reverse transcription polymerase chain reaction. Overexpression of miR-3065-5p also enhanced NHA proliferation, elevated resistance to TMZ, and increased the expression levels of TAA-related markers. In addition, both GSC-EXO-induced and miR-3065-5p-overexpressing NHAs promoted tumorigenesis of GSCs in vivo. Discs Large Homolog 2 (DLG2, downregulated in glioblastoma) is a direct downstream target of miR-3065-5p in TAAs, and DLG2 overexpression could partially reverse the transformation of NHAs into TAAs. Collectively, these data demonstrate that GSC-EXOs induce the transformation of NHAs into TAAs via the miR-3065-5p/DLG2 signaling axis and that TAAs can further promote the tumorigenesis of GSCs. Thus, precisely blocking the interactions between astrocytes and GSCs via exosomes may be a novel strategy to inhibit glioblastoma development, but more in-depth mechanistic studies are still needed.

Acupoint catgut embedding attenuates oxidative stress and cognitive impairment in chronic cerebral ischemia by inhibiting the Ang II/AT1R/NOX axis.

Chronic cerebral ischemia (CCI) is a common neurological disorder, characterized by progressive cognitive impairment. Acupoint catgut embedding (ACE) represents a modern acupuncture form that has shown neuroprotective effects; nevertheless, its effects on CCI and the mechanisms remain largely unknown. Here, we aimed to explore the therapeutic action of ACE in CCI-induced cognitive impairment and its mechanisms. The cognitive function of CCI rats was determined using Morris water maze test, and histopathological changes in the brain were assessed through hematoxylin-eosin (HE) staining. To further explore the molecular mechanisms, the expression levels of oxidative stress markers and the Ang II/AT1R/NOX axis-associated molecules in the hippocampus were evaluated using enzyme-linked immunosorbent assay (ELISA), western blotting, and immunohistochemistry. Here, we observed that ACE treatment alleviated cognitive dysfunction and histopathological injury in CCI rats. Intriguingly, candesartan (an AT1R blocker) enhanced the beneficial effects of ACE on ameliorating cognitive impairment in CCI rats. Mechanistically, ACE treatment blocked the Ang II/AT1R/NOX pathway and subsequently suppressed oxidative stress, thus mitigating cognitive impairment in CCI. Our findings first reveal that ACE treatment could suppress cognitive impairment in CCI, which might be partly due to the suppression of Ang II/AT1R/NOX axis.

Concise and Efficient Synthesis of Sequentially Isomeric Hetero[3]rotaxanes.

Stereoisomerism, stemming from the spatial orientation of components in molecular structures, plays a decisive role in nature. While the unconventional bonding found in mechanically interlocked molecules gives rise to unique expressions of stereochemistry, the exploration of their stereoisomers is still in its infancy. Sequence isomerism, characterized by variations in the ordering of mechanically interlocked components in catenanes and rotaxanes, mirrors the sequence variations found in biological macromolecules. Herein, we report the use of artificial molecular pumps for the precise and simple production of sequentially isomeric hetero[3]rotaxanes. Utilizing redox-driven pumping cassettes with different rings, we have synthesized two hetero[3]rotaxane isomers in high isolated yields from two [2]rotaxanes. This research represents a significant advance in sequential molecular assembly, paving the way for the development of sophisticated, functionalized, mechanically interlocked materials.

Exosomal transfer of activated neutrophil-derived lncRNA CRNDE promotes proliferation and migration of airway smooth muscle cells in asthma.

Activated neutrophil-derived exosomes reportedly contribute to the proliferation of airway smooth muscle cells (ASMCs), thereby aggravating the airway wall remodeling during asthma; however, the specific mechanism remains unclear. Lipopolysaccharide (LPS)-EXO and si-CRNDE-EXO were extracted from the media of human neutrophils treated with LPS and LPS + si-CRNDE (a siRNA targets long non-coding RNA CRNDE), respectively. Human ASMCs were co-cultured with LPS-EXO or si-CRNDE-EXO, and cell viability, proliferation and migration were measured. The interplay of colorectal neoplasia differentially expressed (CRNDE), inhibitor of nuclear factor kappa B kinase subunit beta (IKKß) and nuclear receptor subfamily 2 group C member 2 (TAK1) was explored using RNA immunoprecipitation (RIP) and Co-IP assays. A mouse model of asthma was induced using ovalbumin. CRNDE was upregulated in LPS-EXO and successfully transferred from LPS-treated neutrophils to ASMCs through exosome. Mechanically, CRNDE loaded in LPS-EXO reinforced TAK1-mediated IKKß phosphorylation, thereby activating the nuclear factor kappa B (NF-κB) pathway. Functionally, silencing CRNDE in LPS-EXO, an IKKß inhibitor, and an NF-κB inhibitor all removed the upregulation of cell viability, proliferation and migration induced by LPS-EXO in ASMCs. In the end, the in vivo experiment demonstrated that CRNDE knockdown in neutrophils effectively reduced the thickness of bronchial smooth muscle in a mouse model for asthma. Activated neutrophils-derived CRNDE was transferred to ASMCs through exosomes and activated the NF-κB pathway by enhancing IKKß phosphorylation. The latter promoted the proliferation and migration of ASMCs and then contributed to airway remodeling in asthma.

Prediction of disease-associated nsSNPs by integrating multi-scale ResNet models with deep feature fusion.

More than 6000 human diseases have been recorded to be caused by non-synonymous single nucleotide polymorphisms (nsSNPs). Rapid and accurate prediction of pathogenic nsSNPs can improve our understanding of the principle and design of new drugs, which remains an unresolved challenge. In the present work, a new computational approach, termed MSRes-MutP, is proposed based on ResNet blocks with multi-scale kernel size to predict disease-associated nsSNPs. By feeding the serial concatenation of the extracted four types of features, the performance of MSRes-MutP does not obviously improve. To address this, a second model FFMSRes-MutP is developed, which utilizes deep feature fusion strategy and multi-scale 2D-ResNet and 1D-ResNet blocks to extract relevant two-dimensional features and physicochemical properties. FFMSRes-MutP with the concatenated features achieves a better performance than that with individual features. The performance of FFMSRes-MutP is benchmarked on five different datasets. It achieves the Matthew's correlation coefficient (MCC) of 0.593 and 0.618 on the PredictSNP and MMP datasets, which are 0.101 and 0.210 higher than that of the existing best method PredictSNP1. When tested on the HumDiv and HumVar datasets, it achieves MCC of 0.9605 and 0.9507, and area under curve (AUC) of 0.9796 and 0.9748, which are 0.1747 and 0.2669, 0.0853 and 0.1335, respectively, higher than the existing best methods PolyPhen-2 and FATHMM (weighted). In addition, on blind test using a third-party dataset, FFMSRes-MutP performs as the second-best predictor (with MCC and AUC of 0.5215 and 0.7633, respectively), when compared with the other four predictors. Extensive benchmarking experiments demonstrate that FFMSRes-MutP achieves effective feature fusion and can be explored as a useful approach for predicting disease-associated nsSNPs. The webserver is freely available at http://csbio.njust.edu.cn/bioinf/ffmsresmutp/ for academic use.

MAResNet: predicting transcription factor binding sites by combining multi-scale bottom-up and top-down attention and residual network.

Accurate identification of transcription factor binding sites is of great significance in understanding gene expression, biological development and drug design. Although a variety of methods based on deep-learning models and large-scale data have been developed to predict transcription factor binding sites in DNA sequences, there is room for further improvement in prediction performance. In addition, effective interpretation of deep-learning models is greatly desirable. Here we present MAResNet, a new deep-learning method, for predicting transcription factor binding sites on 690 ChIP-seq datasets. More specifically, MAResNet combines the bottom-up and top-down attention mechanisms and a state-of-the-art feed-forward network (ResNet), which is constructed by stacking attention modules that generate attention-aware features. In particular, the multi-scale attention mechanism is utilized at the first stage to extract rich and representative sequence features. We further discuss the attention-aware features learned from different attention modules in accordance with the changes as the layers go deeper. The features learned by MAResNet are also visualized through the TMAP tool to illustrate that the method can extract the unique characteristics of transcription factor binding sites. The performance of MAResNet is extensively tested on 690 test subsets with an average AUC of 0.927, which is higher than that of the current state-of-the-art methods. Overall, this study provides a new and useful framework for the prediction of transcription factor binding sites by combining the funnel attention modules with the residual network.

MDGF-MCEC: a multi-view dual attention embedding model with cooperative ensemble learning for CircRNA-disease association prediction.

Circular RNA (circRNA) is closely involved in physiological and pathological processes of many diseases. Discovering the associations between circRNAs and diseases is of great significance. Due to the high-cost to verify the circRNA-disease associations by wet-lab experiments, computational approaches for predicting the associations become a promising research direction. In this paper, we propose a method, MDGF-MCEC, based on multi-view dual attention graph convolution network (GCN) with cooperative ensemble learning to predict circRNA-disease associations. First, MDGF-MCEC constructs two disease relation graphs and two circRNA relation graphs based on different similarities. Then, the relation graphs are fed into a multi-view GCN for representation learning. In order to learn high discriminative features, a dual-attention mechanism is introduced to adjust the contribution weights, at both channel level and spatial level, of different features. Based on the learned embedding features of diseases and circRNAs, nine different feature combinations between diseases and circRNAs are treated as new multi-view data. Finally, we construct a multi-view cooperative ensemble classifier to predict the associations between circRNAs and diseases. Experiments conducted on the CircR2Disease database demonstrate that the proposed MDGF-MCEC model achieves a high area under curve of 0.9744 and outperforms the state-of-the-art methods. Promising results are also obtained from experiments on the circ2Disease and circRNADisease databases. Furthermore, the predicted associated circRNAs for hepatocellular carcinoma and gastric cancer are supported by the literature. The code and dataset of this study are available at https://github.com/ABard0/MDGF-MCEC.

PScL-2LSAESM: bioimage-based prediction of protein subcellular localization by integrating heterogeneous features with the two-level SAE-SM and mean ensemble method.

MOTIVATION: Over the past decades, a variety of in silico methods have been developed to predict protein subcellular localization within cells. However, a common and major challenge in the design and development of such methods is how to effectively utilize the heterogeneous feature sets extracted from bioimages. In this regards, limited efforts have been undertaken. RESULTS: We propose a new two-level stacked autoencoder network (termed 2L-SAE-SM) to improve its performance by integrating the heterogeneous feature sets. In particular, in the first level of 2L-SAE-SM, each optimal heterogeneous feature set is fed to train our designed stacked autoencoder network (SAE-SM). All the trained SAE-SMs in the first level can output the decision sets based on their respective optimal heterogeneous feature sets, known as 'intermediate decision' sets. Such intermediate decision sets are then ensembled using the mean ensemble method to generate the 'intermediate feature' set for the second-level SAE-SM. Using the proposed framework, we further develop a novel predictor, referred to as PScL-2LSAESM, to characterize image-based protein subcellular localization. Extensive benchmarking experiments on the latest benchmark training and independent test datasets collected from the human protein atlas databank demonstrate the effectiveness of the proposed 2L-SAE-SM framework for the integration of heterogeneous feature sets. Moreover, performance comparison of the proposed PScL-2LSAESM with current state-of-the-art methods further illustrates that PScL-2LSAESM clearly outperforms the existing state-of-the-art methods for the task of protein subcellular localization. AVAILABILITY AND IMPLEMENTATION: https://github.com/csbio-njust-edu/PScL-2LSAESM. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.