[The long-term efficacy of metformin in megestrol acetate-based fertility-sparing treatment for patients with endometrial atypical hyperplasia and endometrioid endometrial cancer]

Objective: To assess the long-term efficacy of metformin in megestrol acetate (MA)-based fertility-sparing treatment for patients with endometrial atypical hyperplasia (EAH) and endometrioid endometrial cancer (EEC). Methods: The randomized controlled trail study was conducted from October 2013 to October 2017 in the Obstetrics and Gynecology Hospital of Fudan University, Shanghai, China. Patients with EAH or EEC were firstly stratified according to pathology, and randomized to receive MA (160 mg orally, daily) plus metformin (500 mg orally, three times a day) or MA (160 mg orally, daily). Baseline data between two groups of patients were compared. Estimates of time to complete remission (CR) and recurrence-free survival (RFS) were calculated using the Kaplan-Meier method. Cox proportional-hazards regression model was used to estimate hazard ratios (HR) of related factors for recurrence-free survival. Quantitative data were represented by M (Q1, Q3). Results: A total of 150 patients were included, and 76 patients were allocated to receive MA plus metformin with the age of 32.5 (28.0, 36.0), while 74 patients received MA alone with the age of 32.0 (28.0, 36.0). By the end of follow-up period, 96.7% (n=145) of patients achieved complete remission, with a median follow-up time of 57.7 (26.7, 70.5) months. The median CR time for the MA plus metformin group and the MA alone group were 6.3 (3.5, 8.3) months and 6.8 (4.0, 9.3) months, respectively (P=0.193), with 2-year cumulative CR rate of 98.6% and 98.5%, respectively (P=0.879). The median time of RFS was 28.1 (12.5, 57.3) months for the MA plus metformin group and 33.3 (14.1, 62.5) months for the MA alone group (P=0.213), with a cumulative RFS rate of 61.9% and 65.8%, respectively (P=0.560). In the subgroup of non-obese (body mass index<28 kg/m2) patients with EAH, the median RFS times were 25.7 (7.6, 60.3) months and 47.3 (17.5, 64.8) months for the MA plus metformin group and the MA alone group, respectively (P=0.033), with a cumulative RFS rate of 57.5% and 80.6%, respectively (P=0.029). According to Cox proportional hazards regression analysis, undergoing assisted reproductive treatment (HR=2.358, 95%CI: 1.069-5.204, P=0.034) was identified as an independent risk factor for recurrence-free survival after complete remission of endometrial lesions. Conclusion: The long-term follow-up outcome indicates that there is no significant difference in CR time and RFS time between MA plus metformin therapy and MA alone therapy for patients with EAH or EEC.

[Evaluation of the efficacy and safety of intravenous infusion of ferric derisomaltose in the treatment of iron deficiency anemia: a single-center retrospective analysis]

Objective: To investigate the clinical efficacy and safety of ferric derisomaltose injection versus iron sucrose injection in the treatment of iron deficiency anemia (IDA) . Methods: A total of 120 patients with iron deficiency anemia admitted from June 2021 to March 2023 were given intravenous iron supplementation with ferric derisomaltose to assess the efficacy and safety of hemoglobin (HGB) elevation before and after treatment. Simultaneously, the clinical effects of iron supplementation with iron sucrose were compared to those of inpatient patients during the same period. Results: Baseline values were comparable in both groups. Within 12 weeks of treatment, the elevated HGB level in the ferric derisomaltose group was higher than that of the iron sucrose group, with a statistical difference at all time points, and the proportion of HGB increased over 20 g/L in the patients treated for 4 weeks was higher (98.7%, 75.9% ). During the treatment with ferric derisomaltose and iron sucrose, the proportion of mild adverse reactions in the ferric derisomaltose group was slightly lower than that of the iron sucrose group, and neither group experienced any serious adverse reactions. The patients responded well to the infusion treatment, with no reports of pain or pigmentation at the injection site. Conclusion: The treatment of IDA patients with ferric derisomaltose has a satisfactory curative effect, with the advantages of rapidity, accuracy, and safety. Therefore, it is worthy of widespread clinical use.

[Advances in chest imaging in early chronic obstructive pulmonary disease]

Chronic obstructive pulmonary disease(COPD)is a heterogeneous and complex disease, and is characterized by exertional dyspnea and chronic cough. For many years, lung function testing have been used to diagnose COPD, but the sensitivity of lung function testing is low, so there is an urgent need for more sensitive diagnostic methods that show early changes in pathology. In recent years, with the rapid development of HRCT, quantitative CT, new magnetic resonance imaging technology, optical coherence tomography (OCT), artificial intelligence, electrical impedance tomography, etc, it provides a basis for the early diagnosis of COPD. This article reviewed the progress in imaging studies of early COPD in recent years.

A new and effective two-step clustering approach for single cell RNA sequencing data

BACKGROUND

The rapid devolvement of single cell RNA sequencing (scRNA-seq) technology leads to huge amounts of scRNA-seq data, which greatly advance the research of many biomedical fields involving tissue heterogeneity, pathogenesis of disease and drug resistance etc. One major task in scRNA-seq data analysis is to cluster cells in terms of their expression characteristics. Up to now, a number of methods have been proposed to infer cell clusters, yet there is still much space to improve their performance.

RESULTS

In this paper, we develop a new two-step clustering approach to effectively cluster scRNA-seq data, which is called TSC - the abbreviation of Two-Step Clustering. Particularly, by dividing all cells into two types: core cells (those possibly lying around the centers of clusters) and non-core cells (those locating in the boundary areas of clusters), we first clusters the core cells by hierarchical clustering (the first step) and then assigns the non-core cells to the corresponding nearest clusters (the second step). Extensive experiments on 12 real scRNA-seq datasets show that TSC outperforms the state of the art methods.

CONCLUSION

TSC is an effective clustering method due to its two-steps clustering strategy, and it is a useful tool for scRNA-seq data analysis.

HPC-Atlas: Computationally Constructing A Comprehensive Atlas of Human Protein Complexes

A fundamental principle of biology is that proteins tend to form complexes to play important roles in the core functions of cells. For a complete understanding of human cellular functions, it is crucial to have a comprehensive atlas of human protein complexes. Unfortunately, we still lack such a comprehensive atlas of experimentally validated protein complexes, which prevents us from gaining a complete understanding of the compositions and functions of human protein complexes, as well as the underlying biological mechanisms. To fill this gap, we built Human Protein Complexes Atlas (HPC-Atlas), as far as we know, the most accurate and comprehensive atlas of human protein complexes available to date. We integrated two latest protein interaction networks, and developed a novel computational method to identify nearly 9000 protein complexes, including many previously uncharacterized complexes. Compared with the existing methods, our method achieved outstanding performance on both testing and independent datasets. Furthermore, with HPC-Atlas we identified 751 severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-affected human protein complexes, and 456 multifunctional proteins that contain many potential moonlighting proteins. These results suggest that HPC-Atlas can serve as not only a computing framework to effectively identify biologically meaningful protein complexes by integrating multiple protein data sources, but also a valuable resource for exploring new biological findings. The HPC-Atlas webserver is freely available at http://www.yulpan.top/HPC-Atlas.

Self-supervised learning with chemistry-aware fragmentation for effective molecular property prediction

Molecular property prediction (MPP) is a crucial and fundamental task for AI-aided drug discovery (AIDD). Recent studies have shown great promise of applying self-supervised learning (SSL) to producing molecular representations to cope with the widely-concerned data scarcity problem in AIDD. As some specific substructures of molecules play important roles in determining molecular properties, molecular representations learned by deep learning models are expected to attach more importance to such substructures implicitly or explicitly to achieve better predictive performance. However, few SSL pre-trained models for MPP in the literature have ever focused on such substructures. To challenge this situation, this paper presents a Chemistry-Aware Fragmentation for Effective MPP (CAFE-MPP in short) under the self-supervised contrastive learning framework. First, a novel fragment-based molecular graph (FMG) is designed to represent the topological relationship between chemistry-aware substructures that constitute a molecule. Then, with well-designed hard negative pairs, a is pre-trained on fragment-level by contrastive learning to extract representations for the nodes in FMGs. Finally, a Graphormer model is leveraged to produce molecular representations for MPP based on the embeddings of fragments. Experiments on 11 benchmark datasets show that the proposed CAFE-MPP method achieves state-of-the-art performance on 7 of the 11 datasets and the second-best performance on 3 datasets, compared with six remarkable self-supervised methods. Further investigations also demonstrate that CAFE-MPP can learn to embed molecules into representations implicitly containing the information of fragments highly correlated to molecular properties, and can alleviate the over-smoothing problem of graph neural networks.

Bird Object Detection: Dataset Construction, Model Performance Evaluation, and Model Lightweighting

The application of object detection technology has a positive auxiliary role in advancing the intelligence of bird recognition and enhancing the convenience of bird field surveys. However, challenges arise due to the absence of dedicated bird datasets and evaluation benchmarks. To address this, we have not only constructed the largest known bird object detection dataset, but also compared the performances of eight mainstream detection models on bird object detection tasks and proposed feasible approaches for model lightweighting in bird object detection. Our constructed bird detection dataset of GBDD1433-2023, includes 1433 globally common bird species and 148,000 manually annotated bird images. Based on this dataset, two-stage detection models like Faster R-CNN and Cascade R-CNN demonstrated superior performances, achieving a Mean Average Precision (mAP) of 73.7% compared to one-stage models. In addition, compared to one-stage object detection models, two-stage object detection models have a stronger robustness to variations in foreground image scaling and background interference in bird images. On bird counting tasks, the accuracy ranged between 60.8% to 77.2% for up to five birds in an image, but this decreased sharply beyond that count, suggesting limitations of object detection models in multi-bird counting tasks. Finally, we proposed an adaptive localization distillation method for one-stage lightweight object detection models that are suitable for offline deployment, which improved the performance of the relevant models. Overall, our work furnishes an enriched dataset and practice guidelines for selecting suitable bird detection models.

Molecular property prediction by semantic-invariant contrastive learning

MOTIVATION

Contrastive learning has been widely used as pretext tasks for self-supervised pre-trained molecular representation learning models in AI-aided drug design and discovery. However, existing methods that generate molecular views by noise-adding operations for contrastive learning may face the semantic inconsistency problem, which leads to false positive pairs and consequently poor prediction performance.

RESULTS

To address this problem, in this article, we first propose a semantic-invariant view generation method by properly breaking molecular graphs into fragment pairs. Then, we develop a Fragment-based Semantic-Invariant Contrastive Learning (FraSICL) model based on this view generation method for molecular property prediction. The FraSICL model consists of two branches to generate representations of views for contrastive learning, meanwhile a multi-view fusion and an auxiliary similarity loss are introduced to make better use of the information contained in different fragment-pair views. Extensive experiments on various benchmark datasets show that with the least number of pre-training samples, FraSICL can achieve state-of-the-art performance, compared with major existing counterpart models.

AVAILABILITY AND IMPLEMENTATION

The code is publicly available at https://github.com/ZiqiaoZhang/FraSICL.

PCGAN: a generative approach for protein complex identification from protein interaction networks

MOTIVATION

Protein complexes are groups of polypeptide chains linked by non-covalent protein-protein interactions, which play important roles in biological systems and perform numerous functions, including DNA transcription, mRNA translation, and signal transduction. In the past decade, a number of computational methods have been developed to identify protein complexes from protein interaction networks by mining dense subnetworks or subgraphs.

RESULTS

In this article, different from the existing works, we propose a novel approach for this task based on generative adversarial networks, which is called PCGAN, meaning identifying Protein Complexes by GAN. With the help of some real complexes as training samples, our method can learn a model to generate new complexes from a protein interaction network. To effectively support model training and testing, we construct two more comprehensive and reliable protein interaction networks and a larger gold standard complex set by merging existing ones of the same organism (including human and yeast). Extensive comparison studies indicate that our method is superior to existing protein complex identification methods in terms of various performance metrics. Furthermore, functional enrichment analysis shows that the identified complexes are of high biological significance, which indicates that these generated protein complexes are very possibly real complexes.

AVAILABILITY AND IMPLEMENTATION

https://github.com/yul-pan/PCGAN.

[Role and clinical significance of MUC4 gene mutations in thrombotic events in patients with classic paroxysmal nocturnal hemoglobinuria]

Objective: This study aimed to investigate the role and clinical significance of MUC4 gene mutations in thrombotic events in patients with classic paroxysmal nocturnal hemoglobinuria (PNH) patients. Methods: A retrospective analysis was conducted on the clinical data and gene sequencing results of 45 patients with classic PNH admitted to the Department of Hematology, Tianjin Medical University General Hospital, from June 2018 to February 2022. MUC4 gene mutations in patients with classic PNH were summarized, and the risk factors for thrombotic events in these patients were analyzed. Additionally, the effects of MUC4 gene mutations on the cumulative incidence and survival of thrombotic events in patients with classic PNH were determined. Results: The detection rate of MUC4 gene mutations in patients with classic PNH who experienced thrombotic events (thrombotic group) was 68.8% (11/16), which was significantly higher than that in the non-thrombotic group [10.3% (3/29) ] (P<0.001). All mutations occurred in exon 2. MUC4 mutation (OR=20.815, P=0.010) was identified as an independent risk factor for thrombotic events in patients with classic PNH. The cumulative incidence of thrombotic events was 78.6% (11/14) in the MUC4 gene mutation group (mutation group) and 16.1% (5/31) in the non-mutation group, showing a statistically significant difference between the two groups (P<0.001). Survival analysis showed a lower overall survival (OS) rate in the thrombotic group compared with that in the non-thrombotic group [ (34.4±25.2) % vs. (62.7±19.3) % ] (P=0.045). The OS rate of patients was (41.7±29.9) % in the mutation group and (59.1±18.3) % in the non-mutation group (P=0.487) . Conclusion: MUC4 gene mutations are associated with an increased incidence of thrombotic events in classic PNH patients, highlighting their role as independent risk factors for thrombosis in this population. These mutations can be considered a novel predictive factor that aids in evaluating the risk of thrombosis in patients with classic PNH.

Molecular Property Prediction by Contrastive Learning with Attention-Guided Positive Sample Selection

MOTIVATION

Predicting molecular properties is one of the fundamental problems in drug design and discovery. In recent years, self-supervised learning has shown its promising performance in image recognition, natural language processing, and single-cell data analysis. Contrastive learning is a typical self-supervised learning method used to learn the features of data so that the trained model can more effectively distinguish the data. One important issue of contrastive learning is how to select positive samples for each training example, which will significantly impact the performance of contrastive learning.

RESULTS

In this paper, we propose a new method for molecular property prediction by Contrastive  Learning with Attention-guided Positive-sample Selection (CLAPS). Firstly, we generate positive samples for each training example based on an attention-guided selection scheme. Secondly, we employ a Transformer encoder to extract latent feature vectors and compute the contrastive loss aiming to distinguish positive and negative sample pairs. Finally, we use the trained encoder for predicting molecular properties. Experiments on various benchmark datasets show that our approach outperforms the state-of-the-art (SOTA) methods in most cases.

AVAILABILITY

The code is publicly available at https://github.com/wangjx22/CLAPS.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

[A new form of familial platelet disorder caused by germline mutations in RUNX1 in a pedigree]

Objective: To investigate the clinical and biological characteristics of familial platelet disorder (FPD) with germline Runt-related transcription factor (RUNX) 1 mutations. Methods: Patients diagnosed with myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML) with RUNX1 mutations from February 2016 to December 2021 in Wuhan No.1 Hospital underwent pedigree analysis and were screened for gene mutations (somatic and germline). Patients diagnosed with FPD with germline RUNX1 mutations were enrolled and evaluated in terms of clinical characteristics and biological evolution. Bioinformatics analysis was used to assess the pathogenicity of mutations and to analyze the effect of mutated genes on the function of the corresponding protein. Results: Germline RUNX1 mutations were detected in three out of 34 patients suffering from MDS/AML who had RUNX1 mutations. A pedigree of FPD with RUNX1 (RUNX1-FPD) c.562A>C and RUNX1 c.1415T>C mutations was diagnosed, and the mutations were of patrilineal origin. Bioinformatics analysis indicated that the locus at positions 188 and 472 in the AML-1G type of RUNX1 was highly conserved across different species, and that variations might influence functions of the proteins. The mutations were evaluated to be highly pathogenic. Of the nine cases with germline RUNX1 mutations: two patients died due AML progression; one case with AML survived without leukemia after transplantation of hemopoietic stem cells; four patients showed mild-to-moderate thrombocytopenia; two cases had no thrombocytopenia. During the disease course of the proband and her son, mutations in RUNX1, NRAS and/or CEBPA and KIT appeared in succession, and expression of cluster of differentiation-7 on tumor cells was enhanced gradually. None of the gene mutations correlated with the tumor were detected in the four cases not suffering from MDS/AML, and they survived until the end of follow-up. Conclusions: RUNX1-FPD was rare. The mutations c.562A>C and c.1415T>C of RUNX1 could be the disease-causing genes for the family with RUNX1-FPD, and these mutations could promote malignant transformation. Biological monitoring should be carried out regularly to aid early intervention for family members with RUNX1-FPD.

A182 PREDICTING HIGH DIRECT HEALTHCARE COSTS IN PEDIATRIC PATIENTS WITH INFLAMMATORY BOWEL DISEASE IN THE FIRST YEAR FOLLOWING DIAGNOSIS

Background

The incidence of inflammatory bowel disease (IBD) continues to rise rapidly among Canadian children. The care of children results in higher direct healthcare costs than adults with IBD. It is imperative that we identify individuals who will become the highest-cost users of the health system in order to intervene early and decrease the individual- and system-level burden of IBD.

Purpose

To develop a predictive-model for high-cost health system users and (2) identify factors associated with high-cost healthcare use.

Method

Incident cases of IBD diagnosed ≤17y residing in Ontario and enrolled in the Canadian Children IBD Network (CIDsCaNN) between Dec 31 2013 and Jan 31 2019 were linked deterministically using health card number to health administrative data. Using a validated algorithm, direct healthcare costs accumulated between the 31^st and 365^th day after diagnosis were calculated using data from CIDsCaNN (medications) and health administrative data (health system encounters, including surgery). A predictive model was created to determine high-cost (≤25^th percentile) and medium-cost (26^th to 75^th) users, compared to low-cost users (>75th) using ordinal logistic regression. Potential predictive variables were determined a priori based on clinical significance and magnitude of univariable association, based on sample size-informed degrees of freedom. Variables from CIDsCaNN data included: IBD type, age at diagnosis, sex, first line of therapy (steroids, aminosalicylates, exclusive enteral nutrition; yes or no, not mutually exclusive), disease activity (severe vs. moderate vs. none/mild based on PUCAI [UC] or wPCDAI [Crohn’s]). Predictive variables from the health administrative data included: rural/urban residence, hospitalization at diagnosis, intestinal resection or colectomy within 3 months of diagnosis, emergency department visit ±1 month of diagnosis, and a mental health encounter within the first year following diagnosis. Anti-TNF treatment was excluded from models due to the strong correlation with the outcome (direct costs). Overall model fit was estimated with a c-statistic.

Result(s)

Among the 487 (57% Crohn’s) children included in the study, the mean (sd) direct costs accumulated between the 31^st and 365^th days following IBD diagnosis was $14,451 (14,665). The mean cost among high-cost users was $33,533 (16,530); medium-cost users, $11,038 (5322); low-cost users, $2530 (831). The predictive model identified high-cost users of the health system with acceptable model fit (c-statistic 0.69). The relative contribution of individual variables, as measured by odds ratio (OR), is reported in the Table.

Image

Conclusion(s)

The direct healthcare costs of pediatric IBD are substantial. Children with IBD who become high-cost users of the health system were identifiable using characteristics at diagnosis (e.g., need for mental health care, emergency visits, older age). Further research should assess whether interventions in patients at-risk for becoming high-cost users may help to reduce costs.

Please acknowledge all funding agencies by checking the applicable boxes below

Other

Please indicate your source of funding;

Ontario Academic Health Sciences Centres Alternate Funding Plan Innovation Fund

Disclosure of Interest

E. Kuenzig: None Declared, R. Duchen: None Declared, T. Walters Grant / Research support from: Janssen, Abbvie, Psfizer, Ferring, Amgen, Consultant of: Janssen, Abbvie, Psfizer, Ferring, Amgen, D. Mack: None Declared, A. Griffiths Grant / Research support from: Abbvie, Consultant of: Abbvie, Amgen, BristolMyersSquibb, Janssen, Lilly, Takeda, Speakers bureau of: Abbvie, Janssen, Takeda, C. Bernstein Grant / Research support from: Research grants from Abbvie Canada, Amgen Canada, Pfizer Canada, and Sandoz Canada and contract grants from Janssen, Abbvie and Pfizer, Consultant of: Abbvie Canada, Amgen Canada, Bristol Myers Squibb Canada, JAMP Pharmaceuticals, Janssen Canada, Pfizer Canada, Sandoz Canada, Takeda, Speakers bureau of: Abbvie Canada, Janssen Canada, Pfizer Canada and Takeda Canada, G. Kaplan Grant / Research support from: Ferring, Consultant of: AbbVie, Janssen, Pfizer, Amgen, Sandoz, Pendophram, and Takeda, Speakers bureau of: AbbVie, Janssen, Pfizer, Amgen, Sandoz, Pendophram, and Takeda, A. Otley Grant / Research support from: Research support: AbbVie Global. Research site: AbbVie, Pfizer, Eli-Lily, Janssen, Consultant of: AbbVie Canada, W. Yu: None Declared, X. Wang: None Declared, J. Guan: None Declared, S. Fung: None Declared, E. Benchimol Consultant of: McKesson Canada, Dairy Farmers of Ontario (unrelated to medications used to treat inflammatory bowel disease)

Causal Gene Identification Using Non-Linear Regression-Based Independence Tests

With the development of biomedical techniques in the past decades, causal gene identification has become one of the most promising applications in human genome-based business, which can help doctors to evaluate the risk of certain genetic diseases and provide further treatment recommendations for potential patients. When no controlled experiments can be applied, machine learning techniques like causal inference-based methods are generally used to identify causal genes. Unfortunately, most of the existing methods detect disease-related genes by ranking-based strategies or feature selection techniques, which generally return a superset of the corresponding real causal genes. There are also some causal inference-based methods that can identify a part of real causal genes from those supersets, but they are just able to return a few causal genes. This is contrary to our knowledge, as many results from controlled experiments have demonstrated that a certain disease, especially cancer, is usually related to dozens or hundreds of genes. In this work, we present an effective approach for identifying causal genes from gene expression data by using a new search strategy based on non-linear regression-based independence tests, which is able to greatly reduce the search space, and simultaneously establish the causal relationships from the candidate genes to the disease variable. Extensive experiments on real-world cancer datasets show that our method is superior to the existing causal inference-based methods in three aspects: 1) our method can identify dozens of causal genes, and 1/3  ∼ 1/2 of the discovered causal genes can be verified by existing works that they are really directly related to the corresponding disease; 2) The discovered causal genes are able to distinguish the status or disease subtype of the target patient; 3) Most of the discovered causal genes are closely relevant to the disease variable.

Effectively Clustering Single Cell RNA Sequencing Data by Sparse Representation

Clustering analysis has been widely used in analyzing single-cell RNA-sequencing (scRNA-seq) data to study various biological problems at cellular level. Although a number of scRNA-seq data clustering methods have been developed, most of them evaluate the similarity of pairwise cells while ignoring the global relationships among cells, which sometimes cannot effectively capture the latent structure of cells. In this paper, we propose a new clustering method SPARC for scRNA-seq data. The most important feature of SPARC is a novel similarity metric that uses the sparse representation coefficients of each cell in terms of the other cells to measure the relationships among cells. In addition, we develop an outlier detection method to help parameter selection in SPARC. We compare SPARC with nine existing scRNA-seq data clustering methods on twelve real datasets. Experimental results show that SPARC achieves the state of the art performance. By further analyzing the cell similarity data derived from sparse representations, we find that SPARC is much more effective in mining high quality clusters of scRNA-seq data than two traditional similarity metrics. In conclusion, this study provides a new way to effectively cluster scRNA-seq data and achieves more accurate clustering results than the state of art methods.

[Posterior atlanto-axial intraarticular distraction technique as revision surgery to treat atlanto-axial dislocation associated with basilar invagination]

Objective: To examine the effect of posterior atlanto-axial intraarticular distraction technique as revision surgery for failed posterior fossa decompression in patients with basilar invagination(BI) and atlanto-axial dislocation(AAD). Methods: The clinical data of 13 cases of AAD accompanied with BI treated at Department of Neurosurgery, Xuanwu Hospital, Capital Medical University were retrospectively analyzed. There were 3 males and 10 females,aged (42.6±9.5) years (range:30 to 63 years). All cases had assimilation of atlas and once underwent posterior fossa decompression. Anterior tissue was released through posterior approach followed by cage implantation into facet joint and occipital-cervical fixation with cantilever technique. The clinical results were evaluated using Japanese Orthopedic Association scale(JOA) and the main radiological measurements including atlantodental interval (ADI), the distance of odontoid tip above Chamberlain line(DCL),clivus-canal angle(CCA) and the length of syrinx were collected. Paired sample t test was used to compared the data before and after operation. Results: All patients underwent surgery successfully, the mean surgical time was (187.7±47.4) minutes (range from 116 to 261 minutes). Twenty occipital condyle screws, 26 C2 pedicle screws and 3 occipital plates were implanted. Clinical symptoms improved in all patients. Twelve patients had complete reduction of basilar invagination and atlanto-axial dislocation, 1 achieved near completely reduction of basilar invagination. The postoperative ADI, DCL and CCA significantly improved((4.3±1.1) mm vs. (1.8±0.8) mm, (11.7±5.0) mm vs. (6.4±2.8) mm, (142.4±7.9)° vs. (133.3±7.9)°, all P<0.01).There were 5 cases with syringomyelia before surgery, and shrinkage of syrinx was observed 1 week after surgery in all cases. Eight patients achieved bone fusion 3 months after surgery, all patients achieved bone fusion 6 months after surgery. The JOA score increased from 12.8±2.3 before surgery to 14.8±1.3 one year after surgery, with statistically significant difference (t=4.416, P<0.01).No implant failure, spacer subsidence and infection were observed. Conclusion: In cases of failure posterior fossa decompression of basilar invagination and atlanto-axial dislocation, using posterior atlanto-axial intraarticular distraction and cantilever technique with cage implantation could achieve complete reduction and symptomatic relief.

TiC2D: Trajectory Inference From Single-Cell RNA-Seq Data Using Consensus Clustering

Cellular programs often exhibit strong heterogeneity and asynchrony in the timing of program execution. Single-cell RNA-seq technology has provided an unprecedented opportunity for characterizing these cellular processes by simultaneously quantifying many parameters at single-cell resolution. Robust trajectory inference is a critical step in the analysis of dynamic temporal gene expression, which can shed light on the mechanisms of normal development and diseases. Here, we present TiC2D, a novel algorithm for cell trajectory inference from single-cell RNA-seq data, which adopts a consensus clustering strategy to precisely cluster cells. To evaluate the power of TiC2D, we compare it with three state-of-the-art methods on four independent single-cell RNA-seq datasets. The results show that TiC2D can accurately infer developmental trajectories from single-cell transcriptome. Furthermore, the reconstructed trajectories enable us to identify key genes involved in cell fate determination and to obtain new insights about their roles at different developmental stages.

Use of a urinary catheter as a guidewire to facilitate safe nasotracheal intubation following iatrogenic retropharyngeal perforation

Nasotracheal intubation facilitates adequate access for surgical procedures in the oral cavity, is not limited by mouth opening and can be better tolerated by patients in intensive care. Complications of nasotracheal intubation can include epistaxis, turbinate injury and sinusitis. Retropharyngeal submucosal perforation by the tracheal tube has also been infrequently reported. Here, we report a case of difficult nasotracheal intubation resulting in retropharyngeal submucosal perforation in a patient with a history of obstructive sleep apnoea listed for uvulopalatopharyngoplasty. To facilitate successful tracheal re-intubation, we used a soft urinary catheter via the other nostril. In this report, we discuss the reasons why this complication may have occurred, how submucosal perforation could be avoided, and reflect on our management of this rare complication. Difficult nasotracheal intubation can present unique challenges, and airway management plans should be made accordingly.

Effective drug-target interaction prediction with mutual interaction neural network

Motivation

Accurately predicting drug–target interaction (DTI) is a crucial step to drug discovery. Recently, deep learning techniques have been widely used for DTI prediction and achieved significant performance improvement. One challenge in building deep learning models for DTI prediction is how to appropriately represent drugs and targets. Target distance map and molecular graph are low dimensional and informative representations, which however have not been jointly used in DTI prediction. Another challenge is how to effectively model the mutual impact between drugs and targets. Though attention mechanism has been used to capture the one-way impact of targets on drugs or vice versa, the mutual impact between drugs and targets has not yet been explored, which is very important in predicting their interactions.

Results

Therefore, in this article we propose MINN-DTI, a new model for DTI prediction. MINN-DTI combines an interacting-transformer module (called Interformer) with an improved Communicative Message Passing Neural Network (CMPNN) (called Inter-CMPNN) to better capture the two-way impact between drugs and targets, which are represented by molecular graph and distance map, respectively. The proposed method obtains better performance than the state-of-the-art methods on three benchmark datasets: DUD-E, human and BindingDB. MINN-DTI also provides good interpretability by assigning larger weights to the amino acids and atoms that contribute more to the interactions between drugs and targets.

Availability and implementation

The data and code of this study are available at https://github.com/admislf/MINN-DTI.

Learning Causal Structures Based on Divide and Conquer

This article addresses two important issues of causal inference in the high-dimensional situation. One is how to reduce redundant conditional independence (CI) tests, which heavily impact the efficiency and accuracy of existing constraint-based methods. Another is how to construct the true causal graph from a set of Markov equivalence classes returned by these methods. For the first issue, we design a recursive decomposition approach where the original data (a set of variables) are first decomposed into two small subsets, each of which is then recursively decomposed into two smaller subsets until none of these subsets can be decomposed further. Redundant CI tests can be reduced by inferring causalities from these subsets. The advantage of this decomposition scheme lies in two aspects: 1) it requires only low-order CI tests and 2) it does not violate d -separation. The complete causality can be reconstructed by merging all the partial results of the subsets. For the second issue, we employ regression-based CI tests to check CIs in linear non-Gaussian additive noise cases, which can identify more causal directions by [Formula: see text] (or [Formula: see text]). Consequently, causal direction learning is no longer limited by the number of returned V -structures and consistent propagation. Extensive experiments show that the proposed method can not only substantially reduce redundant CI tests but also effectively distinguish the equivalence classes.

[Sleep quality and sleep disturbances in Chinese pregnant women: a multicenter cross-sectional study]

Objective: This study aims to investigate the sleep quality of pregnant women in Xuhui District, Shanghai, and the related factors of sleep disturbances during pregnancy. Methods: From February 2019 to February 2021, we used online integrated sleep questionnaire (including PSQI, BQ, ESS, AIS) in Shanghai Jiao Tong University School of Medicine Affiliated Sixth People's Hospital, The International Peace Maternity and Child Health Hospitals of China Welfare Institution, and Shanghai Eighth People's Hospital, to investigate the sleep quality across pregnancy. We also collected maternal physical examination results, childbearing history, sociodemographic, and other clinical data. The prevalences and related factors of various sleep disturbances in pregnant women were analyzed, including insufficient/excessive nighttime sleep, low sleep efficiency, difficulty falling asleep, poor sleep quality, insomnia, daytime sleepiness, and high risk of sleep-disordered breathing (SDB). Results: This study includes 1 898 cases in the first trimester (T1), 3 099 cases in the second trimester (T2), and 1 539 cases in the third trimester (T3). Poor sleep quality (38.6%), daytime sleepiness (mild 41.9%, moderate 17.7%, severe 2.1%), and suspicious insomnia (32.3%) are most prevalent among women in T1 (P<0.01). In comparison, short sleep time (2.7%), long sleep time (8.6%), difficulty falling asleep (12.2%), poor sleep efficiency (35.4%), very poor sleep quality (6.7%), clinical insomnia (21.8%), and high-risk SDB (6.4%) are most prevalent among women in T3 (P<0.05). During pregnancy, late gestation (OR=1.016, 95%CI: 1.006-1.025) and multiple induced/drug abortions (OR=1.329, 95%CI: 1.043-1.692) are risk factors for poor sleep quality (PSQI>5), while multiple full-term deliveries (OR=0.800, 95%CI: 0.675-0.949) is its protective factor. Advanced maternal age (OR=0.976, 95%CI: 0.956-0.997), multiple full-term deliveries (OR=0.808, 95%CI: 0.680-0.959), late gestation (OR=0.983, 95%CI: 0.974-0.992) and hypertension (OR=0.572, 95%CI: 0.401-0.814) are protective factors for daytime sleepiness (ESS>6). The high-risk pregnancy category (OR=9.312, 95%CI: 1.156-74.978) is a risk factor for insomnia (AIS≥4), while multiple full-term deliveries (OR=0.815, 95%CI: 0.691-0.961) is its protective factor. High BMI (OR=1.334, 95%CI: 1.270-1.402) and hypertension (OR=4.427, 95%CI: 2.539-7.719) are risk factors for high-risk SDB in pregnant women. Conclusions: The prevalences of various sleep disturbances are high throughout pregnancy. Noticeably, symptoms of maternal SDB develop along with pregnancy. Different types of sleep disturbances are associated with different factors. Women of high-risk pregnancy category, in late gestation, with high BMI, hypertension, a history of induced/drug abortion, or without a history of full-term delivery can be at high risk of sleep disturbances during pregnancy.

Link Weight Prediction Using Weight Perturbation and Latent Factor

Link weight prediction is an important subject in network science and machine learning. Its applications to social network analysis, network modeling, and bioinformatics are ubiquitous. Although this subject has attracted considerable attention recently, the performance and interpretability of existing prediction models have not been well balanced. This article focuses on an unsupervised mixed strategy for link weight prediction. Here, the target attribute is the link weight, which represents the correlation or strength of the interaction between a pair of nodes. The input of the model is the weighted adjacency matrix without any preprocessing, as widely adopted in the existing models. Extensive observations on a large number of networks show that the new scheme is competitive to the state-of-the-art algorithms concerning both root-mean-square error and Pearson correlation coefficient metrics. Analytic and simulation results suggest that combining the weight consistency of the network and the link weight-associated latent factors of the nodes is a very effective way to solve the link weight prediction problem.

Deep structural clustering for single-cell RNA-seq data jointly through autoencoder and graph neural network

Single-cell RNA sequencing (scRNA-seq) permits researchers to study the complex mechanisms of cell heterogeneity and diversity. Unsupervised clustering is of central importance for the analysis of the scRNA-seq data, as it can be used to identify putative cell types. However, due to noise impacts, high dimensionality and pervasive dropout events, clustering analysis of scRNA-seq data remains a computational challenge. Here, we propose a new deep structural clustering method for scRNA-seq data, named scDSC, which integrate the structural information into deep clustering of single cells. The proposed scDSC consists of a Zero-Inflated Negative Binomial (ZINB) model-based autoencoder, a graph neural network (GNN) module and a mutual-supervised module. To learn the data representation from the sparse and zero-inflated scRNA-seq data, we add a ZINB model to the basic autoencoder. The GNN module is introduced to capture the structural information among cells. By joining the ZINB-based autoencoder with the GNN module, the model transfers the data representation learned by autoencoder to the corresponding GNN layer. Furthermore, we adopt a mutual supervised strategy to unify these two different deep neural architectures and to guide the clustering task. Extensive experimental results on six real scRNA-seq datasets demonstrate that scDSC outperforms state-of-the-art methods in terms of clustering accuracy and scalability. Our method scDSC is implemented in Python using the Pytorch machine-learning library, and it is freely available at https://github.com/DHUDBlab/scDSC.

[Clinical and genetic spectrum of SCN2A gene associated epilepsy and episodic ataxia]

Objective: To explore the clinical manifestations and genetic characteristics of patients with epilepsy and episodic ataxia caused by SCN2A gene variation. Methods: The clinical data of seizure manifestation, imaging examination and genetic results of 5 patients with epilepsy and (or) episodic ataxia because of SCN2A gene variation admitted to the Department of Pediatrics, the Third Affiliated Hospital of Zhengzhou University from July 2017 to January 2021 were analyzed retrospectively. Results: Among 5 patients, 4 were female and 1 was male. The onset age of epilepsy ranged from 4 days to 8 months. There were 2 cases of benign neonatal or infantile epilepsy and 3 cases of epileptic encephalopathy, in whom 1 case had development retardation,1 case transformed from West syndrome to infantile spasm and another one transformed from infantile spasm to Lennox-Gastaut syndrome. One case of benign neonatal-infantile epilepsy was characterized by neonatal onset seizures and episodic ataxia developed at the age of 78 months. Electroencephalograms at first visit of 5 cases showed that 2 cases were normal, 1 case had focal epileptic discharge, and 2 cases had multi-focal abnormal discharge with peak arrhythmia. The brain magnetic resonance imaging (MRI) of 3 cases were nomal, 1 case was abnormal (brain atrophy with decreased white matter) and the results of 1 case was unknown. The follow-up time ranged from 17 months to 89 months. Four cases of epilepsy were controlled and 1 case died at 2 years of age. Two cases had normal intelligence and motor development, 2 had moderate to severe intelligence retardation and motor critical state, and 1 had moderate to severe intelligence and motor development retardation. SCN2A gene variations were identified in all cases. There were 4 missense variations and 1 frameshift variation. Three variations had not been reported so far, including c.4906A>G,c.3643G>T,c.638delT. Conclusions: Variations in SCN2A gene can cause benign neonatal or infantile epilepsy and epileptic encephalopathy. Some children develop episodic ataxia with growing age. The variation of SCN2A gene is mainly missense variation.

Protein Function Prediction Based on PPI Networks: Network Reconstruction vs Edge Enrichment

Over the past decades, massive amounts of protein-protein interaction (PPI) data have been accumulated due to the advancement of high-throughput technologies, and but data quality issues (noise or incompleteness) of PPI have been still affecting protein function prediction accuracy based on PPI networks. Although two main strategies of network reconstruction and edge enrichment have been reported on the effectiveness of boosting the prediction performance in numerous literature studies, there still lack comparative studies of the performance differences between network reconstruction and edge enrichment. Inspired by the question, this study first uses three protein similarity metrics (local, global and sequence) for network reconstruction and edge enrichment in PPI networks, and then evaluates the performance differences of network reconstruction, edge enrichment and the original networks on two real PPI datasets. The experimental results demonstrate that edge enrichment work better than both network reconstruction and original networks. Moreover, for the edge enrichment of PPI networks, the sequence similarity outperformes both local and global similarity. In summary, our study can help biologists select suitable pre-processing schemes and achieve better protein function prediction for PPI networks.

[Obstructive sleep apnea and metabolic syndrome: an association study based on a large sample clinical database]

Objective: To investigate the prevalence and associated risk factors of metabolic syndrome (MS) in patients with obstructive sleep apnea (OSA). Methods: From July 2007 to June 2017, a total of 8 155 adult subjects, including 6 484 males and 1 671 females, aged 18-90 (43.13±12.28), body mass index 14.61~59.56 (25.59±3.98) kg/m²,who were admitted to the Department of Otorhinolaryngology head and Neck surgery of The Sixth People's Hospital affiliated to Shanghai Jiao Tong University, were retrospectively analyzed. All patients underwent polysomnography and biochemical tests. Subjects were divided into four groups (non-OSA, mild OSA, moderate OSA, and severe OSA) according to OSA severity. The prevalence of MS was expressed as percentage, and the correlation between OSA and MS and its characteristic pathophysiological indicators was evaluated by logistic regression model after adjusting for factors such as gender, age, BMI, neck circumference, hip circumference, smoking and alcohol consumption, and was expressed by odds ratio (OR). SPSS 25.0 software was used for statistical analysis. Results: The overall prevalence of MS was 43.6%, and that of non-/mild/moderate/severe OSA group was 18.6%, 30.4%, 43.8%, 57.1%.Logistic regression showed that patients with mild/moderate/severe OSA had an increased risk of MS compared with non-OSA patients, with adjusted OR values and confidence intervals of 1.27 (1.05-1.54), 1.84 (1.53-2.22), and 2.08 (1.76-2.46), respectively (P<0.01).In addition, indicators of OSA anoxic burden [oxygen drop index(T_oxygen=7.1), minimum blood oxygen(T_minimum=56.3), blood oxygen saturation below 90% cumulative time ratio (T_CT90=10.6) ]were closely associated with MS disease(P<0.01), but sleep fragmentation index (arousals index) was not significantly associated with MS disease. Conclusion: The risk of MS gradually increases with the severity of OSA, and the indicators reflecting OSA hypoxia burden are closely related to MS disease.

Protein-protein interaction prediction based on ordinal regression and recurrent convolutional neural networks

Background

Protein protein interactions (PPIs) are essential to most of the biological processes. The prediction of PPIs is beneficial to the understanding of protein functions and thus is helpful to pathological analysis, disease diagnosis and drug design etc. As the amount of protein data is growing fast in the post genomic era, high-throughput experimental methods are expensive and time-consuming for the prediction of PPIs. Thus, computational methods have attracted researcher’s attention in recent years. A large number of computational methods have been proposed based on different protein sequence encoders.

Results

Notably, the confidence score of a protein sequence pair could be regarded as a kind of measurement to PPIs. The higher the confidence score for one protein pair is, the more likely the protein pair interacts. Thus in this paper, a deep learning framework, called ordinal regression and recurrent convolutional neural network (OR-RCNN) method, is introduced to predict PPIs from the perspective of confidence score. It mainly contains two parts: the encoder part of protein sequence pair and the prediction part of PPIs by confidence score. In the first part, two recurrent convolutional neural networks (RCNNs) with shared parameters are applied to construct two protein sequence embedding vectors, which can automatically extract robust local features and sequential information from the protein pairs. Based on it, the two embedding vectors are encoded into one novel embedding vector by element-wise multiplication. By taking the ordinal information behind confidence score into consideration, ordinal regression is used to construct multiple sub-classifiers in the second part. The results of multiple sub-classifiers are aggregated to obtain the final confidence score. Following that, the existence of PPIs is determined by the confidence score. We set a threshold \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\theta$$\end{document}θ, and say the interaction exists between the protein pair if its confidence score is bigger than \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\theta$$\end{document}θ.

Conclusions

We applied our method to predict PPIs on data sets S. cerevisiae and Homo sapiens. Through experimental verification, our method outperforms state-of-the-art PPI prediction models.

FraGAT: a fragment-oriented multi-scale graph attention model for molecular property prediction

MOTIVATION

Molecular property prediction is a hot topic in recent years. Existing graph-based models ignore the hierarchical structures of molecules. According to the knowledge of chemistry and pharmacy, the functional groups of molecules are closely related to its physio-chemical properties and binding affinities. So, it should be helpful to represent molecular graphs by fragments that contain functional groups for molecular property prediction.

RESULTS

In this article, to boost the performance of molecule property prediction, we first propose a definition of molecule graph fragments that may be or contain functional groups, which are relevant to molecular properties, then develop a fragment-oriented multi-scale graph attention network for molecular property prediction, which is called FraGAT. Experiments on several widely used benchmarks are conducted to evaluate FraGAT. Experimental results show that FraGAT achieves state-of-the-art predictive performance in most cases. Furthermore, our case studies show that when the fragments used to represent the molecule graphs contain functional groups, the model can make better predictions. This conforms to our expectation and demonstrates the interpretability of the proposed model.

AVAILABILITY AND IMPLEMENTATION

The code and data underlying this work are available in GitHub, at https://github.com/ZiqiaoZhang/FraGAT.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

[The effect of diallyl sulfide on the lung tissue of rats poisoned by paraquat]

Objective: To investigate the mechanism of diallyl sulfide (DAS) on paraquat (PQ) - induced acute lung injury in rats. Methods: In May 2016, 32 adult male Wistar rats were randomly divided into control group, model (PQ) group, DAS treatment group and dexamethasone (DXM) treatment group, with 8 rats in each group. PQ poisoning model was established by intragastric administration of PQ solution (70 mg/kg) . 100 mg/kg DAS (DAS treatment group) , normal saline (control group and PQ group) and 1 mg/kg DXM (DXM treatment group) were injected intraperitoneally before and after modeling. After 24 hours, the rats were killed and the degree of lung injury was observed. The expression of inducible nitric oxide synthase (iNOS) in lung tissue was measured. Alveolar macrophages were isolated and cultured. The supernatant was taken to determine the content of NO, and the expressions of iNOS mRNA in alveolar macrophages were detected. Results: Compared with the control group, the pathological injury score and the expression of iNOS in the lung tissue of PQ group were significantly increased, and the content of NO secreted by alveolar macrophages and the expression of iNOS mRNA were significantly increased (P<0.05) . Compared with PQ group, the pathological injury scores and the expressions of iNOS in lung tissue of rats in DAS treatment group and DXM treatment group were significantly decreased, and the contents of NO secreted by alveolar macrophages and the expressions of iNOS mRNA were significantly decreased (P<0.05) . There was no significant difference between DXM group and DAS group (P>0.05) . Conclusion: DAS may have protective effect on acute lung injury induced by PQ in rats.

Denoising Protein-Protein interaction network via variational graph auto-encoder for protein complex detection

Identifying protein complexes is an important issue in computational biology, as it benefits the understanding of cellular functions and the design of drugs. In the past decades, many computational methods have been proposed by mining dense subgraphs in Protein-Protein Interaction Networks (PINs). However, the high rate of false positive/negative interactions in PINs prevents accurately detecting complexes directly from the raw PINs. In this paper, we propose a denoising approach for protein complex detection by using variational graph auto-encoder. First, we embed a PIN to vector space by a stacked graph convolutional network (GCN), then decide which interactions in the PIN are credible. If the probability of an interaction being credible is less than a threshold, we delete the interaction. In such a way, we reconstruct a reliable PIN. Following that, we detect protein complexes in the reconstructed PIN by using several typical detection methods, including CPM, Coach, DPClus, GraphEntropy, IPCA and MCODE, and compare the results with those obtained directly from the original PIN. We conduct the empirical evaluation on four yeast PPI datasets (Gavin, Krogan, DIP and Wiphi) and two human PPI datasets (Reactome and Reactomekb), against two yeast complex benchmarks (CYC2008 and MIPS) and three human complex benchmarks (REACT, REACT_uniprotkb and CORE_COMPLEX_human), respectively. Experimental results show that with the reconstructed PINs obtained by our denoising approach, complex detection performance can get obviously boosted, in most cases by over 5%, sometimes even by 200%. Furthermore, we compare our approach with two existing denoising methods (RWS and RedNemo) while varying different matching rates on separate complex distributions. Our results show that in most cases (over 2/3), the proposed approach outperforms the existing methods.

Boosting scRNA-seq data clustering by cluster-aware feature weighting

Background

The rapid development of single-cell RNA sequencing (scRNA-seq) enables the exploration of cell heterogeneity, which is usually done by scRNA-seq data clustering. The essence of scRNA-seq data clustering is to group cells by measuring the similarities among genes/transcripts of cells. And the selection of features for cell similarity evaluation is of great importance, which will significantly impact clustering effectiveness and efficiency.

Results

In this paper, we propose a novel method called CaFew to select genes based on cluster-aware feature weighting. By optimizing the clustering objective function, CaFew obtains a feature weight matrix, which is further used for feature selection. The genes have large weights in at least one cluster or the genes whose weights vary greatly in different clusters are selected. Experiments on 8 real scRNA-seq datasets show that CaFew can obviously improve the clustering performance of existing scRNA-seq data clustering methods. Particularly, the combination of CaFew with SC3 achieves the state-of-art performance. Furthermore, CaFew also benefits the visualization of scRNA-seq data.

Conclusion

CaFew is an effective scRNA-seq data clustering method due to its gene selection mechanism based on cluster-aware feature weighting, and it is a useful tool for scRNA-seq data analysis.

DEVELOPING twin-specific 75-g oral glucose tolerance test diagnostic thresholds for gestational diabetes based on the risk of future maternal diabetes: a population-based cohort study

OBJECTIVE

To develop twin-specific outcome-based oral glucose tolerance test (OGTT) diagnostic thresholds for GDM based on the risk of future maternal type-2 diabetes.

DESIGN

A population-based retrospective cohort study (2007-2017).

SETTING

Ontario, Canada.

METHODS

Nulliparous women with a live singleton (n = 55 361) or twin (n = 1308) birth who underwent testing for gestational diabetes mellitus (GDM) using a 75-g OGTT in Ontario, Canada (2007-2017). We identified the 75-g OGTT thresholds in twin pregnancies that were associated with similar incidence rates of future type-2 diabetes to those associated with the standard OGTT thresholds in singleton pregnancies.

RESULTS

For any given 75-g OGTT value, the incidence rate of future maternal type-2 diabetes was lower for women with a twin than women with a singleton pregnancy. Using women with a negative OGTT as reference, the risk of future maternal type-2 diabetes in twin pregnancies with a positive OGTT based on the standard OGTT thresholds (9.86 per 1000 person years, adjusted hazard ratio (aHR) 4.79, 95% CI 2.69-8.51) was lower than for singleton pregnancies with a positive OGTT (18.74 per 1000 person years, aHR 8.22, 95% CI 7.38-9.16). The twin-specific OGTT fasting, 1-hour and 2-hour thresholds identified in the current study based on correlation with future maternal type-2 diabetes were 5.8 mmol/l (104 mg/dl), 11.8 mmol/l (213 mg/dl) and 10.4 mmol/l (187 mg/dl), respectively.

CONCLUSIONS

We identified potential twin-specific OGTT thresholds for GDM that are associated with a similar risk of future type-2 diabetes to that observed in women diagnosed with GDM in singleton pregnancies based on standard OGTT thresholds.

TWEETABLE ABSTRACT

Potential twin-specific OGTT thresholds for GDM were identified.

Classifying Cognitive Normal and Early Mild Cognitive Impairment of Alzheimer’s Disease by Applying Restricted Boltzmann Machine to fMRI Data

Background:

Neuroimaging is an important tool in early detection of Alzheimer’s disease (AD), which is a serious neurodegenerative brain disease among the elderly subjects. Independent component analysis (ICA) is arguably one of the most widely used algorithm for the analysis of brain imaging data, which can be used to extract intrinsic networks of brain from functional magnetic resonance imaging (fMRI).

Method:

Witnessed by recent studies, a more flexible model known as restricted Boltzmann machine (RBM) can also be used to extract spatial maps and time courses of intrinsic networks from resting state fMRI, moreover, RBM shows superior temporal features than ICA. Here, we seek to employ RBM to improve the performance of classifying individuals. Experiments are performed on healthy controls and subjects at the early stage of AD, i.e., cognitive normal (CN) and early mild cognitive impairment participants (EMCI), and two types of data, i.e., structural magnetic resonance imaging (sMRI) and fMRI data.

Results:

(1) By separately employing ICA for sMRI and fMRI, the features extracted from fMRI improve classification accuracy by 7.5% for CN and EMCI; (2) instead of applying ICA to fMRI, using RBM further improves classification accuracy by 7.75% for CN and EMCI; (3) the lesions at the early stage of AD are more likely to occur in the regions around slices 4, 6, 10, 14, 19, 51 and 59 of the whole brain in the longitudinal direction.

Conclusion:

By using fMRI instead of sMRI and RBM instead of ICA, we can classify CN and EMCI more efficiently.

Therapeutic efficacy of zoledronic acid combined with calcitriol in elderly patients receiving total hip arthroplasty or hemiarthroplasty for osteoporotic femoral neck fracture

Zoledronic acid could improve the clinical outcome in elderly patients receiving total hip arthroplasty or hemiarthroplasty for osteoporotic femoral neck fracture in the 1-year prospective study.

INTRODUCTION

To validate the therapeutic efficacy of zoledronic acid (ZOL) in elderly patients with femoral neck fracture who received total hip arthroplasty (THA) or hemiarthroplasty (HA).

METHODS

Included in this study were 95 elderly patients with femoral neck fractures who received THA/HA between August 2015 and June 2018. They were randomized into a ZOL group and a control group. Patients in ZOL group received a yearly single dose of 5 mg ZOL intravenous injection plus 0.5 μg/day calcitriol and 1000 mg/day calcium carbonate 2 days before THA or HA. Patients in the control group were treated with the same dose of calcitriol and calcium carbonate only without ZOL. Bone mineral density (BMD) was measured by dual-energy X-ray absorptiometry. Bone metabolism markers including the total extension of the peptide type I collagen amino end (P1NP) and beta collagen degradation product (β-CTX) were obtained by serum examination. The postoperative functional outcome was assessed using Harris Hip Score (HHS).

RESULTS

During the follow-up period, BMD in the ZOL group was improved and significantly higher than that in the control group at 6 and 12 months post-operation. Bone metabolism markers P1NP and β-CTX in ZOL group remained at a relatively low level as compared with that in the control group at 6 months after treatment. No significant difference in the mean HHS and the excellent/good rate of joint function was observed during the follow-up period between the two groups. The occurrence of adverse events in the ZOL group was significantly higher than that in the control group.

CONCLUSIONS

A single infusion of ZOL shows promise in improving BMD of the healthy side of the femoral neck, lumbar spine, and total hip and decreasing the level of bone markers, which may improve the clinical outcome of patients with osteoporotic femoral neck fractures receiving THA/HA.

[Identification of a novel mutation of SOX10 gene and analysis of the phenotype]

Objective: To explore the clinical features and pathogenic mechanisms of a special syndrome with congenital sensorineural hearing loss, albinism, heterochromia iridis, nystagmus and myelin dysplasia. Methods: Detailed medical history, systematic audiology tests, ophthalmic and neurological examinations were carried out to analyze the clinical features of the child, and further molecular genetic tests including chromosome karyotype analysis, and deafness gene screening were conducted. Results: A new de novo heterozygous mutation (c.336G>T/p.Met112Ile) was detected in the child, while both his parents were demonstrated to be wild-type and symptom free. The analysis of clinical features indicated the diagnosis of PCW syndrome. Conclusion: This study identified a new mutation of SOX10 gene, which enriched the mutation spectrum of this gene. And the analysis of clinical characteristics of this patient also expanded the phenotype of this gene. This study provided a reference for clinical diagnosis and genetic diagnosis of PCW syndrome.

Generation and propagation of hyperbolic secant solitons, Peregrine solitons, and breathers in a coherently prepared atomic system

We study the generation and propagation of hyperbolic secant solitons, Peregrine solitons, and various breathers in a coherently prepared three-level atomic system, where two lower states are coherently prepared prior to the injection of a strong pump field and a weak probe field. We show that a flat dispersion without gain and loss along with a large Kerr nonlinearity can be achieved in a broad range of probe field frequency. Moreover, optical hyperbolic secant solitons can be easily achieved in such a broad range at a very low light intensity and propagate stably. Due to the enhanced Kerr nonlinearity, we also show that it is possible to generate optical rogue waves and breathers with very weak light stimulus, which is three orders of magnitude smaller than that used in nonlinear fibers. Because the gain/absorption is negligible and the quantum noise of the probe field can be significantly suppressed, our work may pave the way for realizing solitons, rogue waves, and breathers at the quantum level.

Association between indoor air pollution, tobacco smoke and tuberculosis: an updated systematic review and meta-analysis

OBJECTIVE

This study aims to further quantify evidence of the association between exposure to indoor air pollution (IAP), tobacco smoke etc., on the one hand and the risk of contracting tuberculosis (TB) on the other.

STUDY DESIGN

This was a systematic review and meta-analysis of articles published between June 2014 and February 2020 in PubMed, Web of Science, among others.

METHODS

We only included studies that controlled for confounders, screened both the exposed and unexposed study participants, and passive smoking studies that limited the study population to non-smokers. Quality of studies was assessed using the Newcastle-Ottawa scale. The analysis was conducted using STATA, and pooled effect sizes were calculated using the random-effects model, and heterogeneity was tested for using the Cochran Q test and I2 statistic.

RESULTS

A total of 26 articles were included in the final analysis. There was an increased risk of contracting TB among people exposed to IAP (risk ratio [RR] = 1.68, 95% confidence interval [CI] 1.108-2.542). We also observed a two-fold increase in the risk of contracting TB from exposure to secondhand tobacco smoke (RR = 2.15, 95%CI 1.419-3.242). Tobacco smoking doubled the risk of contracting TB (RR = 2.67, 95%CI 2.017-3.527). Furthermore, studies that used microbiological tests showed a higher RR compared to those that used other TB diagnostic methods.

CONCLUSION

Exposure to IAP and secondhand tobacco smoke increases the risk of contracting TB. Various disease prevention campaigns should include IAP awareness and encourage a shift to cleaner sources of energy.

Computationally identifying hot spots in protein-DNA binding interfaces using an ensemble approach

BACKGROUND

Protein-DNA interaction governs a large number of cellular processes, and it can be altered by a small fraction of interface residues, i.e., the so-called hot spots, which account for most of the interface binding free energy. Accurate prediction of hot spots is critical to understand the principle of protein-DNA interactions. There are already some computational methods that can accurately and efficiently predict a large number of hot residues. However, the insufficiency of experimentally validated hot-spot residues in protein-DNA complexes and the low diversity of the employed features limit the performance of existing methods.

RESULTS

Here, we report a new computational method for effectively predicting hot spots in protein-DNA binding interfaces. This method, called PreHots (the abbreviation of Predicting Hotspots), adopts an ensemble stacking classifier that integrates different machine learning classifiers to generate a robust model with 19 features selected by a sequential backward feature selection algorithm. To this end, we constructed two new and reliable datasets (one benchmark for model training and one independent dataset for validation), which totally consist of 123 hot spots and 137 non-hot spots from 89 protein-DNA complexes. The data were manually collected from the literature and existing databases with a strict process of redundancy removal. Our method achieves a sensitivity of 0.813 and an AUC score of 0.868 in 10-fold cross-validation on the benchmark dataset, and a sensitivity of 0.818 and an AUC score of 0.820 on the independent test dataset. The results show that our approach outperforms the existing ones.

CONCLUSIONS

PreHots, which is based on stack ensemble of boosting algorithms, can reliably predict hot spots at the protein-DNA binding interface on a large scale. Compared with the existing methods, PreHots can achieve better prediction performance. Both the webserver of PreHots and the datasets are freely available at: http://dmb.tongji.edu.cn/tools/PreHots/ .