Ab-Initio Membrane Protein Amphipathic Helix Structure Prediction Using Deep Neural Networks.

Amphipathic helix (AH)features the segregation of polar and nonpolar residues and plays important roles in many membrane-associated biological processes through interacting with both the lipid and the soluble phases. Although the AH structure has been discovered for a long time, few ab initio machine learning-based prediction models have been reported, due to the limited amount of training data. In this study, we report a new deep learning-based prediction model, which is composed of a residual neural network and the uneven-thresholds decision algorithm. It is constructed on 121 membrane proteins, in total 51640 residue samples, which are curated from an up-to-date membrane protein structure database. Through a rigid 10-fold nested cross-validation experiment, we demonstrate that our model can achieve promising predictions and exceed current state-of-the-art approaches in this field. This presents a new avenue for accurately predicting AHs. Analysis on the contribution of the input residues and some cases further reveals the high interpretability and the generalization of our model.

CoCoPRED: coiled-coil protein structural feature prediction from amino acid sequence using deep neural networks.

MOTIVATION: Coiled-coil is composed of two or more helices that are wound around each other. It widely exists in proteins and has been discovered to play a variety of critical roles in biology processes. Generally, there are three types of structural features in coiled-coil: coiled-coil domain (CCD), oligomeric state and register. However, most of the existing computational tools only focus on one of them. RESULTS: Here, we describe a new deep learning model, CoCoPRED, which is based on convolutional layers, bidirectional long short-term memory, and attention mechanism. It has three networks, i.e. CCD network, oligomeric state network, and register network, corresponding to the three types of structural features in coiled-coil. This means CoCoPRED has the ability of fulfilling comprehensive prediction for coiled-coil proteins. Through the 5-fold cross-validation experiment, we demonstrate that CoCoPRED can achieve better performance than the state-of-the-art models on both CCD prediction and oligomeric state prediction. Further analysis suggests the CCD prediction may be a performance indicator of the oligomeric state prediction in CoCoPRED. The attention heads in CoCoPRED indicate that registers a, b and e are more crucial for the oligomeric state prediction. AVAILABILITY AND IMPLEMENTATION: CoCoPRED is available at http://www.csbio.sjtu.edu.cn/bioinf/CoCoPRED. The datasets used in this research can also be downloaded from the website. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

GraphBind: protein structural context embedded rules learned by hierarchical graph neural networks for recognizing nucleic-acid-binding residues.

Knowledge of the interactions between proteins and nucleic acids is the basis of understanding various biological activities and designing new drugs. How to accurately identify the nucleic-acid-binding residues remains a challenging task. In this paper, we propose an accurate predictor, GraphBind, for identifying nucleic-acid-binding residues on proteins based on an end-to-end graph neural network. Considering that binding sites often behave in highly conservative patterns on local tertiary structures, we first construct graphs based on the structural contexts of target residues and their spatial neighborhood. Then, hierarchical graph neural networks (HGNNs) are used to embed the latent local patterns of structural and bio-physicochemical characteristics for binding residue recognition. We comprehensively evaluate GraphBind on DNA/RNA benchmark datasets. The results demonstrate the superior performance of GraphBind than state-of-the-art methods. Moreover, GraphBind is extended to other ligand-binding residue prediction to verify its generalization capability. Web server of GraphBind is freely available at http://www.csbio.sjtu.edu.cn/bioinf/GraphBind/.

Protein-ligand binding residue prediction enhancement through hybrid deep heterogeneous learning of sequence and structure data.

MOTIVATION: Knowledge of protein-ligand binding residues is important for understanding the functions of proteins and their interaction mechanisms. From experimentally solved protein structures, how to accurately identify its potential binding sites of a specific ligand on the protein is still a challenging problem. Compared with structure-alignment-based methods, machine learning algorithms provide an alternative flexible solution which is less dependent on annotated homogeneous protein structures. Several factors are important for an efficient protein-ligand prediction model, e.g. discriminative feature representation and effective learning architecture to deal with both the large-scale and severely imbalanced data. RESULTS: In this study, we propose a novel deep-learning-based method called DELIA for protein-ligand binding residue prediction. In DELIA, a hybrid deep neural network is designed to integrate 1D sequence-based features with 2D structure-based amino acid distance matrices. To overcome the problem of severe data imbalance between the binding and nonbinding residues, strategies of oversampling in mini-batch, random undersampling and stacking ensemble are designed to enhance the model. Experimental results on five benchmark datasets demonstrate the effectiveness of proposed DELIA pipeline. AVAILABILITY AND IMPLEMENTATION: The web server of DELIA is available at www.csbio.sjtu.edu.cn/bioinf/delia/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Recent methodology progress of deep learning for RNA-protein interaction prediction.

Interactions between RNAs and proteins play essential roles in many important biological processes. Benefitting from the advances of next generation sequencing technologies, hundreds of RNA-binding proteins (RBP) and their associated RNAs have been revealed, which enables the large-scale prediction of RNA-protein interactions using machine learning methods. Till now, a wide range of computational tools and pipelines have been developed, including deep learning models, which have achieved remarkable performance on the identification of RNA-protein binding affinities and sites. In this review, we provide an overview of the successful implementation of various deep learning approaches for predicting RNA-protein interactions, mainly focusing on the prediction of RNA-protein interaction pairs and RBP-binding sites on RNAs. Furthermore, we discuss the advantages and disadvantages of these approaches, and highlight future perspectives on how to design better deep learning models. Finally, we suggest some promising future directions of computational tasks in the study of RNA-protein interactions, especially the interactions between noncoding RNAs and proteins. This article is categorized under: RNA Interactions with Proteins and Other Molecules > Protein-RNA Interactions: Functional Implications RNA Evolution and Genomics > Computational Analyses of RNA RNA Interactions with Proteins and Other Molecules > Protein-RNA Recognition.

A Self-Training Subspace Clustering Algorithm under Low-Rank Representation for Cancer Classification on Gene Expression Data.

Accurate identification of the cancer types is essential to cancer diagnoses and treatments. Since cancer tissue and normal tissue have different gene expression, gene expression data can be used as an efficient feature source for cancer classification. However, accurate cancer classification directly using original gene expression profiles remains challenging due to the intrinsic high-dimension feature and the small size of the data samples. We proposed a new self-training subspace clustering algorithm under low-rank representation, called SSC-LRR, for cancer classification on gene expression data. Low-rank representation (LRR) is first applied to extract discriminative features from the high-dimensional gene expression data; the self-training subspace clustering (SSC) method is then used to generate the cancer classification predictions. The SSC-LRR was tested on two separate benchmark datasets in control with four state-of-the-art classification methods. It generated cancer classification predictions with an overall accuracy 89.7 percent and a general correlation 0.920, which are 18.9 and 24.4 percent higher than that of the best control method respectively. In addition, several genes (RNF114, HLA-DRB5, USP9Y, and PTPN20) were identified by SSC-LRR as new cancer identifiers that deserve further clinical investigation. Overall, the study demonstrated a new sensitive avenue to recognize cancer classifications from large-scale gene expression data.

Protein tyrosine phosphatases receptor type D is a potential tumour suppressor gene inactivated by deoxyribonucleic acid methylation in paediatric acute myeloid leukaemia.

AIM: Protein tyrosine phosphatases receptor type D (PTPRD) is a tumour suppressor gene, and its epigenetic silencing is frequently found in glioblastoma. As aberrant deoxyribonucleic acid (DNA) methylation patterning has been shown to play a role in leukaemogenesis, we studied the promoter methylation, expression profiles and molecular functions of PTPRD in paediatric patients with acute myeloid leukaemia (AML). METHODS: Bone marrow specimens were obtained from 32 Chinese patients with a mean age of 7.2 years (range 1.1-16.5). PTPRD and methylation status were evaluated by real-time polymerase chain reaction (PCR) and methylation-specific PCR. Western blot and flow cytometry techniques were also used. RESULTS: PTPRD expression was decreased by promoter region methylation in six AML cells and methylated in 21 (65.6%) of the 32 samples. In addition, PTPRD expression could be induced by the DNA demethylating agent 5-aza-2'-deoxycytidine. Furthermore, functional studies showed that overexpression of PTPRD in AML cells inhibited cell proliferation and clonogenicity as well as inducing apoptosis. However, PTPRD knockdown increased cell proliferation. These effects were associated with downregulation of cyclin D1, c-myc and upregulation of Bax. CONCLUSION: The results of this study demonstrated that PTPRD was a potential tumour suppressor gene inactivated by DNA methylation in paediatric AML.

Comparison of thoracolaparoscopic esophagectomy with cervical anastomosis with McKeown esophagectomy for middle esophageal cancer.

BACKGROUND: In China, the middle esophageal squamous cell cancer is the most common tumor type, and Mckeown esophagectomy (ME) is preferably adopted by thoracic surgeon. But, the surgical trauma of ME is great. Thoracolaparoscopic esophagectomy (TE) was developed to decrease the operative stress; however, the safety and efficacy were not defined. In this study, clinical outcomes were compared between patients who received ME and TE. METHODS: The data of 113 patients who suffered from middle-thoracic esophageal cancer during the same period were collected. Sixty-two patients received ME (ME group), and 51 patients received TE (TE group). Patients' demographics and short-term clinicopathologic outcomes were comparable between the two groups. Survival rate was estimated using the Kaplan-Meier method, and comparisons between groups were performed with log-rank test. RESULTS: Patients in TE group had lower body mass index (BMI). Preoperative tumor stage in TE group was much earlier. Both overall and thoracic operation time were longer in TE group. The blood loss during operation and postoperative day (POD) 1 was less in TE group, which contributed to the less blood transfusion. In TE group, postoperative incidence of pulmonary complications and atrial fibrillation (p = 0.035 and p = 0.033) was lower; the inflammatory response and incision pain were significantly alleviated; the ICU and in-hospital stay was shorter as well because of less surgical trauma. No statistically significant difference was found between two groups in terms of overall survival or disease-free survival. CONCLUSIONS: The efficacy and safety of TE were supported by the selected patients in this cohort study. Although it is lack of randomness in this research, some advantages of TE were gratifying such as lower postoperative complications and similar survival with ME. A multicenter prospective randomized study is now required.

Clinical Outcome of Middle Thoracic Esophageal Cancer with Intrathoracic or Cervical Anastomosis.

BACKGROUNDS: What is the optimal way for the middle esophageal cancer? It is still controversial. In this study, the clinical outcome of middle thoracic esophageal cancer with either intrathoracic or cervical anastomosis was analyzed in our department. PATIENTS AND METHODS: A total of 205 patients who suffered from middle thoracic esophageal cancer were divided into two groups. In group A, 91 patients received intrathoracic anastomosis above aortic arch after esophageal resection via single left thoracotomy, and in group B, 114 patients received cervical anastomosis after esophageal resection via right thoracotomy and median laparotomy. Data of these patients were collected, and morbidity and mortality were analyzed retrospectively. Survival rate was estimated using the Kaplan-Meier method and comparisons between groups were performed with log-rank test. Univariate and multivariate analyses were performed using Cox model to look for independent predictors of survival. RESULTS: Postoperative complications occurred more frequently in group B, such as hemorrhage (p = 0.011), wound infection (p = 0.032), and temporary paresis of the recurrent laryngeal nerve (p = 0.001). Morbidity of anastomotic leak was higher in group B (8.8 vs. 2.2%; p = 0.048), but the associated mortality was not increased. The extent of radical esophagectomy and lymphadenectomy was much greater in group B; therefore, longer esophagus was resected that reduced the cancer residual rate, and more positive lymph nodes were detected that enhanced the accuracy of clinical staging. Fortunately, more patients received adjuvant therapy after operation in group B, and the 5-year survival rate was improved. CONCLUSION: Anastomotic leak rate was higher in cervical anastomosis but with lower mortality. The 5-year survival rate was improved in cervical anastomosis group. The present data support the assumption that cervical anastomosis is a safer and more beneficial procedure for patients with middle thoracic esophageal cancer.