ProSTAGE: Predicting Effects of Mutations on Protein Stability by Using Protein Embeddings and Graph Convolutional Networks.

Protein thermodynamic stability is essential to clarify the relationships among structure, function, and interaction. Therefore, developing a faster and more accurate method to predict the impact of the mutations on protein stability is helpful for protein design and understanding the phenotypic variation. Recent studies have shown that protein embedding will be particularly powerful at modeling sequence information with context dependence, such as subcellular localization, variant effect, and secondary structure prediction. Herein, we introduce a novel method, ProSTAGE, which is a deep learning method that fuses structure and sequence embedding to predict protein stability changes upon single point mutations. Our model combines graph-based techniques and language models to predict stability changes. Moreover, ProSTAGE is trained on a larger data set, which is almost twice as large as the most used S2648 data set. It consistently outperforms all existing state-of-the-art methods on mutation-affected problems as benchmarked on several independent data sets. The protein embedding as the prediction input achieves better results than the previous results, which shows the potential of protein language models in predicting the effect of mutations on proteins. ProSTAGE is implemented as a user-friendly web server.

Bidirectional extracellular electron transfer pathways of Geobacter sulfurreducens biofilms: Molecular insights into extracellular polymeric substances.

The basis for bioelectrochemical technology is the capability of electroactive bacteria (EAB) to perform bidirectional extracellular electron transfer (EET) with electrodes, i.e. outward- and inward-EET. Extracellular polymeric substances (EPS) surrounding EAB are the necessary media for EET, but the biochemical and molecular analysis of EPS of Geobacter biofilms on electrode surface is largely lacked. This study constructed Geobacter sulfurreducens-biofilms performing bidirectional EET to explore the bidirectional EET mechanisms through EPS characterization using electrochemical, spectroscopic fingerprinting and proteomic techniques. Results showed that the inward-EET required extracellular redox proteins with lower formal potentials relative to outward-EET. Comparing to the EPS extracted from anodic biofilm (A-EPS), the EPS extracted from cathodic biofilm (C-EPS) exhibited a lower redox activity, mainly due to a decrease of protein/polysaccharide ratio and α-helix content of proteins. Furthermore, less cytochromes and more tyrosine- and tryptophan-protein like substances were detected in C-EPS than in A-EPS, indicating a diminished role of cytochromes and a possible role of other redox proteins in inward-EET. Proteomic analysis identified a variety of redox proteins including cytochrome, iron-sulfur clusters-containing protein, flavoprotein and hydrogenase in EPS, which might serve as an extracellular redox network for bidirectional EET. Those redox proteins that were significantly stimulated in A-EPS and C-EPS might be essential for outward- and inward-EET and warranted further research. This work sheds light on the mechanism of bidirectional EET of G. sulfurreducens biofilms and has implications in improving the performance of bioelectrochemical technology.

Characterization of two novel Fe(III)-reducing and electrogenic bacteria, Shewanella ferrihydritica sp. nov. and Shewanella electrica sp. nov., isolated from mangrove sediment.

Three novel strains in the genus Shewanella, designated A3AT, C31T and C32, were isolated from mangrove sediment samples. They were facultative anaerobic, Gram-stain-negative, rod-shaped, flagellum-harbouring, oxidase- and catalase-positive, electrogenic and capable of using Fe(III) as an electron acceptor during anaerobic growth. Results of phylogenetic analysis based on 16S rRNA gene and genomic sequences revealed that the strains should be assigned to the genus Shewanella. The 16S rRNA gene similarity, average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH) values between the isolates and their closely related species were below the respective cut-off values for species differentiation. The 16S rRNA gene similarity, ANI and dDDH values between strains C31T and C32 were 99.7, 99.9 and 99.9 %, respectively, indicating that they should belong to the same genospecies. Based on polyphasic taxonomic approach, two novel species are proposed, Shewanella ferrihydritica sp. nov. with type strain A3AT (GDMCC 1.2732T=JCM 34899T) and Shewanella electrica sp. nov. with type strain C31T (GDMCC 1.2736T=JCM 34902T).

Prediction of hot spots towards drug discovery by protein sequence embedding with 1D convolutional neural network.

Protein hotspot residues are key sites that mediate protein-protein interactions. Accurate identification of these residues is essential for understanding the mechanism from protein to function and for designing drug targets. Current research has mostly focused on using machine learning methods to predict hot spots from known interface residues, which artificially extract the corresponding features of amino acid residues from sequence, structure, evolution, energy, and other information to train and test machine learning models. The process is cumbersome, time-consuming and laborious to some extent. This paper proposes a novel idea that develops a pre-trained protein sequence embedding model combined with a one-dimensional convolutional neural network, called Embed-1dCNN, to predict protein hotspot residues. In order to obtain large data samples, this work integrates and extracts data from the datasets of ASEdb, BID, SKEMPI and dbMPIKT to generate a new dataset, and adopts the SMOTE algorithm to expand positive samples to form the training set. The experimental results show that the method achieves an F1 score of 0.82 on the test set. Compared with other hot spot prediction methods, our model achieved better prediction performance.

Unveiling targeted cell-free DNA methylation regions through paired methylome analysis of tumor and normal tissues.

Liquid biopsy analysis of cell-free DNA (cfDNA) has revolutionized cancer research by enabling non-invasive assessment of tumor-derived genetic and epigenetic changes. In this study, we conducted a comprehensive paired-sample differential methylation analysis (psDMR) on reprocessed methylation data from two large datasets, CPTAC and TCGA, to identify and validate differentially methylated regions (DMRs) as potential cfDNA biomarkers for head and neck squamous cell carcinoma (HNSC). Our hypothesis is that the paired sample test provides a more suitable and powerful approach for the analysis of heterogeneous cancers like HNSC. The psDMR analysis revealed a significant number of overlapped hypermethylated DMRs between two datasets, indicating the reliability and relevance of these regions for cfDNA methylation biomarker discovery. We identified several candidate genes, including CALCA, ALX4, and HOXD9, which have been previously established as liquid biopsy methylation biomarkers in various cancer types. Furthermore, we demonstrated the efficacy of targeted region analysis using cfDNA methylation data from oral cavity squamous cell carcinoma and nasopharyngeal carcinoma patients, further validating the utility of psDMR analysis in prioritizing cfDNA methylation biomarkers. Overall, our study contributes to the development of cfDNA-based approaches for early cancer detection and monitoring, expanding our understanding of the epigenetic landscape of HNSC, and providing valuable insights for liquid biopsy biomarker discovery not only in HNSC and other cancer types.

Prediction-oriented prognostic biomarker discovery with survival machine learning methods.

Identifying novel and reliable prognostic biomarkers for predicting patient survival outcomes is essential for deciding personalized treatment strategies for diseases such as cancer. Numerous feature selection techniques have been proposed to address the high-dimensional problem in constructing prediction models. Not only does feature selection lower the data dimension, but it also improves the prediction accuracy of the resulted models by mitigating overfitting. The performances of these feature selection methods when applied to survival models, on the other hand, deserve further investigation. In this paper, we construct and compare a series of prediction-oriented biomarker selection frameworks by leveraging recent machine learning algorithms, including random survival forests, extreme gradient boosting, light gradient boosting and deep learning-based survival models. Additionally, we adapt the recently proposed prediction-oriented marker selection (PROMISE) to a survival model (PROMISE-Cox) as a benchmark approach. Our simulation studies indicate that boosting-based approaches tend to provide superior accuracy with better true positive rate and false positive rate in more complicated scenarios. For demonstration purpose, we applied the proposed biomarker selection strategies to identify prognostic biomarkers in different modalities of head and neck cancer data.

Echoes in a "tree hole": A qualitative study on the service needs and responses at an antidrug helpline in China.

INTRODUCTION: China's antidrug measures have been slowly shifting from police-intervention and punitive approaches to supportive services. However, the system is still highly stigmatizing. Helpline services emerged to engage drug users, families, and friends and provide needed support as they seek rehabilitation. This study aimed to explore service needs expressed during helpline calls, operators' use of techniques when responding to different needs, and operators' experiences working at and views toward the helpline. METHODS: We conducted a qualitative mixed-method study using two sources of data. One source was 47 call recordings collected at a drug helpline in China, and the other was five individual and two focus group interviews conducted with 18 helpline operators. Using a six-step thematic analysis method, we explored the patterns of needs expression and response, and the operators' experiences of interacting with callers. RESULTS: We found that typical callers were drug users and their relatives or friends. Interactions between the callers and operators involved the expression of and response to needs that emerged from involvement with drugs. Informational and emotional needs were the most common. Operators would respond to these needs with different counseling techniques, such as providing information, advising, normalizing, focusing, and instilling hope. The operators developed a system of practices, such as internal supervision, case summaries, and listening back, to enhance competence and ensure quality of services. The helpline work also prompted their critical reflections on the current antidrug system and gradually reshaped their views toward the population they serve. CONCLUSIONS: Antidrug workers who engaged in answering helpline calls employed varying techniques to facilitate callers' expressed needs. They helped by providing much-needed informational and emotional support for drug users, families, and friends. Helpline services opened a private channel for people involved in drug use to express their needs and seek formal help in China's still stigmatizing and punitive antidrug system. Experiences working with anonymous help-seekers outside the statutory rehabilitation system helped workers at the helpline to gain unique reflective insight into the antidrug system and drug users.

Statistical and Machine Learning Methods for Discovering Prognostic Biomarkers for Survival Outcomes.

Discovering molecular biomarkers for predicting patient survival outcomes is an essential step toward improving prognosis and therapeutic decision-making in the treatment of severe diseases such as cancer. Due to the high-dimensionality nature of omics datasets, statistical methods such as the least absolute shrinkage and selection operator (Lasso) have been widely applied for cancer biomarker discovery. Due to their scalability and demonstrated prediction performance, machine learning methods such as XGBoost and neural network models have also been gaining popularity in the community recently. However, compared to more traditional survival methods such as Kaplan-Meier and Cox regression methods, high-dimensional methods for survival outcomes are still less well known to biomedical researchers. In this chapter, we will discuss the key analytical procedures in employing these methods for identifying biomarkers associated with survival data. We will also identify important considerations that emerged from the analysis of actual omics data. Some typical instances of misapplication and misinterpretation of machine learning methods will also be discussed. Using lung cancer and head and neck cancer datasets as demonstrations, we provide step-by-step instructions and sample R codes for prioritizing prognostic biomarkers.

Xin-Li-Fang efficacy and safety for patients with chronic heart failure: A study protocol for a randomized, double-blind, and placebo-controlled trial.

Introduction: Xin-Li-Fang (XLF), a representative Chinese patent medicine, was derived from years of clinical experience by academician Chen Keji, and is widely used to treat chronic heart failure (CHF). However, there remains a lack of high-quality evidence to support clinical decision-making. Therefore, we designed a randomized controlled trial (RCT) to evaluate the efficacy and safety of XLF for CHF. Methods and design: This multicenter, double-blinded RCT will be conducted in China. 300 eligible participants will be randomly assigned to either an XLF group or a control group at a 1:1 ratio. Participants in the XLF group will receive XLF granules plus routine care, while those in the control group will receive placebo granules plus routine care. The study period is 26 weeks, including a 2-week run-in period, a 12-week treatment period, and a 12-week follow-up. The primary outcome is the proportion of patients whose serum NT-proBNP decreased by more than 30%. The secondary outcomes include quality of life, the NYHA classification evaluation, 6-min walking test, TCM symptom evaluations, echocardiography parameters, and clinical events (including hospitalization for worsening heart failure, all-cause death, and other major cardiovascular events). Discussion: The results of the study are expected to provide evidence of high methodological and reporting quality on the efficacy and safety of XLF for CHF. Clinical trial registration: Chinese Clinical Trial Registration Center (www.chictr.org.cn). The trial was registered on 13 April 2022 (ChiCTR2200058649).

A subnetwork-based framework for prioritizing and evaluating prognostic gene modules from cancer transcriptome data.

Cancer prognosis prediction is critical to the clinical decision-making process. Currently, the high availability of transcriptome datasets allows us to extract the gene modules with promising prognostic values. However, the biomarker identification is greatly challenged by tumor and patient heterogeneity. In this study, a framework of three subnetwork-based strategies is presented, incorporating hypothesis-driven, data-driven, and literature-based methods with informative visualization to prioritize candidate genes. By applying the proposed approaches to a head and neck squamous cell cancer (HNSCC) transcriptome dataset, we successfully identified multiple HNSCC-specific gene modules with improved prognostic values and mechanism information compared with the standard gene panel selection methods. The proposed framework is general and can be applied to any type of omics data. Overall, the study demonstrates and supports the use of the subnetwork-based approach for distilling reliable and biologically meaningful prognostic factors.

LOX-1 Regulation in Anti-atherosclerosis of Active Compounds of Herbal Medicine: Current Knowledge and the New Insight.

Lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1) have recently been identified to be closely related to the occurrence and development of atherosclerosis (AS). A growing body of evidence has suggested Chinese medicine takes unique advantages in preventing and treating AS. In this review, the related research progress of AS and LOX-1 has been summarized. And the anti-AS effects of 10 active components of herbal medicine through LOX-1 regulation have been further reviewed. As a potential biomarker and target for intervention in AS, LOX-1 targeted therapy might provide a promising and novel approach to atherosclerotic prevention and treatment.

Shewanella shenzhenensis sp. nov., a novel Fe(III)-reducing bacterium with abundant possible cytochrome genes, isolated from mangrove sediment.

A facultative anaerobic bacterium, designated as A25T, was isolated from a mangrove sediment sample collected in Shenzhen, China. Cells of strain A25T were found to be Gram-staining negative, rod-shaped, flagella-harboring, and oxidase- and catalase-positive. The isolate was able to grow at 4-40 °C (optimum 28 °C) and pH 5.0-9.0 (optimum pH 6.0), and in 0-10% NaCl concentration (w/v) (optimum 1%). Strain A25T was capable of reducing Fe(III) citrate under anaerobic conditions. The major fatty acids of this strain was C16:1ω7c/C16:1ω6c (summed feature 3), C17:1ω8c and iso-C15:0. Results of phylogenetic analyses based on 16S rRNA gene sequences indicated that strain A25T is affiliated with the genus Shewanella, showing the highest similarity to Shewanella seohaensis S7-3T (98.4% similarity). The average nucleotide identity and digital DNA-DNA hybridization values between the genomes of strain A25T and its closely related strains were ≤ 79.0% and ≤ 22.8%, respectively. Based on its phenotypic, phylogenetic properties and physiological and biochemical characteristics, strain A25T (= JCM 34900T = GDMCC 1.2731T) was designated as the type strain of a novel species of the genus Shewanella, for which the name Shewanella shenzhenensis sp. nov. was proposed.

Efficient gradient boosting for prognostic biomarker discovery.

MOTIVATION: A gradient boosting decision tree (GBDT) is a powerful ensemble machine-learning method that has the potential to accelerate biomarker discovery from high-dimensional molecular data. Recent algorithmic advances, such as extreme gradient boosting (XGB) and light gradient boosting (LGB), have rendered the GBDT training more efficient, scalable and accurate. However, these modern techniques have not yet been widely adopted in discovering biomarkers for censored survival outcomes, which are key clinical outcomes or endpoints in cancer studies. RESULTS: In this paper, we present a new R package 'Xsurv' as an integrated solution that applies two modern GBDT training frameworks namely, XGB and LGB, for the modeling of right-censored survival outcomes. Based on our simulations, we benchmark the new approaches against traditional methods including the stepwise Cox regression model and the original gradient boosting function implemented in the package 'gbm'. We also demonstrate the application of Xsurv in analyzing a melanoma methylation dataset. Together, these results suggest that Xsurv is a useful and computationally viable tool for screening a large number of prognostic candidate biomarkers, which may facilitate future translational and clinical research. AVAILABILITY AND IMPLEMENTATION: 'Xsurv' is freely available as an R package at: https://github.com/topycyao/Xsurv. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

A two-step ensemble learning for predicting protein hot spot residues from whole protein sequence.

Protein hot spot residues are functional sites in protein-protein interactions. Biological experimental methods are traditionally used to identify hot spot residues, which is laborious and time-consuming. Thus a variety of computational methods were widely used in recent years. Despite the success of computational methods in hot spot identification, most of them are impractical in reality because they can recognize hot spot residues only from known protein-protein interface residues. Therefore, identifying hot spots from whole protein sequence is a meaningful and interesting issue. However, it will bring extreme imbalance between positive and negative samples. Hot spot residues only account for about 1-2% of whole protein sequences. To address the issue, this paper proposes a two-step ensemble model for identifying hot spot residues from extremely unbalanced data set. The model is composed of 134 classifiers constructed by base KNN and SVM. Compared to the previous methods, our model yields good performance with an F1 score of 0.593 on the BID test set. Furthermore, to validate the robustness of our model, it was tested on other three independent test sets and also achieved good predictions. More importantly, the performance of our model tested on unbalanced data set is comparable with other methods tested on balanced hot spot data set.

Mangrovimonas futianensis sp. nov., a novel species isolated from mangrove sediment.

Three bacterial strains, designated as AS18T, AS27 and AS39, were obtained from mangrove sediment sampled in Futian district, Shenzhen, PR China. Cells of these strains were Gram-negative rods with no flagella. They were able to grow at 10-42 °C (optimum, 37 °C), at pH 5-9 (optimum, pH 6) and in 1-11 % (w/v) NaCl (optimum, 2 %). Phylogenetic analysis based on 16S rRNA gene sequences indicated that the new isolates were clustered within the genus Mangrovimonas, closely related to Mangrovimonas yunxiaonensis (95.1 % similarity) and Mangrovimonas spongiae (94.7 % similarity). Phylogenomic analysis based on multiple core genes revealed that the three strains were located in a different cluster from other closely related strains of the genus Mangrovimonas. Digital DNA-DNA hybridization, average nucleotide identity and average amino acid identity values calculated from genome sequences between isolates and type strains were lower than 25, 75 and 72 %, respectively. The dominant fatty acids were iso-C15 : 0 and iso-C15 : 1 G. The main respiratory quinone was identified as MK-6. The major polar lipids contained phosphatidylethanolamine, two unidentified aminolipids and five unidentified lipids. The results of multiphase taxonomy suggested that the three strains should be assigned to a novel species of the genus Mangrovimonas, for which the name Mangrovimonas futianensis sp. nov. is proposed, with the type strain AS18T (=GDMCC 1.2739T=JCM 34871T).

Paradesulfitobacterium ferrireducens gen. nov., sp. nov., a Fe(III)-reducing bacterium from petroleum-contaminated soil and reclassification of Desulfitobacterium aromaticivorans as Paradesulfitobacterium aromaticivorans comb. nov.

A strictly anaerobic bacterium, strain PLL0T, was isolated from petroleum-contaminated soil sampled in Gansu Province, PR China. Cells were rods, non-motile and Gram-stain-positive. The strain grew at 25-37 °C (optimum, 30 °C) in the presence of 0-3 % (w/v) NaCl (optimum, 2 %). Strain PLL0T was able to reduce ferrihydrite, Fe(III) citrate and thiosulphate. The 16S rRNA gene analysis revealed that this strain clustered with the genus Desulfitobacterium, and showed highest similarity to Desulfitobacterium aromaticivorans UKTLT (95.4 %) followed by Desulfitobacterium chlororespirans Co23T (93.9 %). However, strains PLL0T and UKTLT showed no more than 94.0 % similarity to other species of the genus Desulfitobacterium, and formed an independent group in the phylogenetic tree. The average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH) values between strain PLL0T and Desulfitobacterium species (except for D. aromaticivorans) were 67.4-68.5 % and 12.6-12.7 %, respectively, which are far below the threshold for delineation of a new species. Based on ANI, dDDH, average amino acid identity, phylogenetic analysis and physiologic differences from the previously described taxa, we suggest that strain PLL0T represents a novel species of a novel genus, for which the name Paradesulfitobacterium ferrireducens gen. nov. sp. nov. is proposed. The type strain is PLL0T (=MCCC 1K05549=KCTC 25248). We also propose the reclassification of D. aromaticivorans as Paradesulfitobacterium aromaticivorans comb. nov.

Clinicopathological and genomic features of breast mucinous carcinoma.

INTRODUCTION: Mucinous carcinoma (MC) of the breast is a special histological type of breast cancer. Clinicopathological characteristics and genomic features of MC is not fully understood. MATERIALS AND METHODS: 186,497 primary breast cancer patients from SEER database diagnosed with invasive ductal carcinoma (IDC) or MC were included. 801 primary IDC or MC patients from TCGA cohort were included for transcriptomic and genomic analysis. RESULTS: MC patients were older, had lower tumor grade and T and N stage, higher hormone receptor positive proportions and lower HER2 positive proportions than IDC patients. Kaplan-Meier plots showed that the breast cancer-specific survival (BCSS) of MC patients was significantly better than IDC patients (P < 0.001). However, after adjusting for clinicopathological factors, survival advantage of MC disappeared. In terms of genomic features of MC, representative upregulated genes of MC in transcriptomic level were MUC2, TFF1 and CARTPT. Upregulated pathways of MC included neurotransmitter-related pathways. Moreover, MC was featured by the amplification of 6p25.2, 6q12 and 11q12.3. CONCLUSION: MC is a distinct histological subtype compared with IDC in terms of clinicopathological characteristics and genomic features. Further investigation need to be conducted to explore the formation of this specific histological subtype.