Retention in care and antiretroviral therapy adherence among Medicaid beneficiaries with HIV, 2001-2015.

Disparities in HIV care by socioeconomic status, place of residence, and race/ethnicity prevent progress toward epidemic control. No study has comprehensively characterized the HIV care cascade among people with HIV enrolled in Medicaid - an insurance source for low-income individuals in the US. We analyzed data from 246,127 people with HIV enrolled in Medicaid 2001-2015, aged 18-64, living in 14 US states. We estimated monthly prevalence of four steps of the care cascade: retained in care/adherent to ART; retained/not adherent; not retained/adherent; not retained/not adherent. Beneficiaries were retained in care if they had an outpatient care encounter every six months. Adherence was based on medication possession ratio. We estimated prevalence using a non-parametric multi-state approach, accounting for death as a competing event and for Medicaid disenrollment using inverse probability of censoring weights. Across 2001-2015, the proportion of beneficiaries with HIV who were retained/ART adherent increased, overall and in all subgroups. By 2015, approximately half of beneficiaries were retained in care, and 42% of beneficiaries were ART adherent. We saw meaningful differences by race/ethnicity and region. Our work highlights an important disparity in the HIV care cascade by insurance status during this time period.

Incidence of Colon Cancer Among Medicaid Beneficiaries With or Without Human Immunodeficiency Virus Under Comparable Colorectal Cancer Screening Patterns.

Background: People with human immunodeficiency virus (HIV; PWH) in the United States have a lower incidence of colon cancer than the general population. The lower incidence may be explained by differences in receipt of screening. Thus, we sought to estimate colon cancer incidence under scenarios in which Medicaid beneficiaries, with or without HIV, followed the same screening protocols. Methods: We used data from 1.5 million Medicaid beneficiaries who were enrolled in 14 US states in 2001-2015 and aged 50-64 years; 72 747 beneficiaries had HIV. We estimated risks of colon cancer and death by age, censoring beneficiaries when they deviated from 3 screening protocols, which were based on Medicaid's coverage policy for endoscopies during the time period, with endoscopy once every 2, 4, or 10 years. We used inverse probability weights to control for baseline and time-varying confounding and informative loss to follow-up. Analyses were performed overall, by sex, and by race/ethnicity. Results: PWH had a lower incidence of colon cancer than beneficiaries without HIV. Compared with beneficiaries without HIV, the risk difference at age 65 years was -1.6% lower (95% confidence interval, -2.3% to -.7%) among PWH with the 2-year protocol and -0.8% lower (-1.3% to -.3%) with the 10-year protocol. Results were consistent across subgroup and sensitivity analyses. Conclusions: Our findings suggest that the lower risk of colon cancer that has been observed among PWH aged 50-64 years compared with those without HIV is not due to differences in receipt of lower endoscopy. Keywords: colon cancer, colorectal cancer screening, endoscopy, Medicaid, human immunodeficiency virus.

Field evaluation of a simple and rapid diagnostic test, RLDT to detect Shigella and enterotoxigenic E. coli in Indian children.

The diagnostic assays currently used to detect Shigella spp. (Shigella) and enterotoxigenic Escherichia coli (ETEC) are complex or elaborate which make them difficult to apply in resource poor settings where these diseases are endemic. The simple and rapid nucleic acid amplification-based assay "Rapid LAMP-based Diagnostic Test (RLDT)" was evaluated to detect Shigella spp (Shigella) and enterotoxigenic Escherichia coli (ETEC) and determine the epidemiology of these pathogens in Kolkata, India. Stool samples (n = 405) from children under five years old with diarrhea seeking care at the hospitals were tested, and 85(21%) and 68(17%) by RLDT, 91(23%) and 58(14%) by quantitative PCR (qPCR) and 35(9%) and 15(4%) by culture, were positive for Shigella and ETEC, respectively. The RLDT showed almost perfect agreement with qPCR, Kappa 0.96 and 0.89; sensitivity 93% and 98%; specificity 100% and 97% for Shigella and ETEC, respectively. While RLDT detected additional 12% Shigella and 13% ETEC than culture, all culture positives for Shigella and ETEC except one each were also positive by the RLDT, sensitivity 97% and 93% respectively. RLDT is a simple, sensitive, and rapid assay that could be implemented with minimum training in the endemic regions to strengthen the disease surveillance system and rapid outbreak detection.

CycleTrans: Learning Neutral Yet Discriminative Features via Cycle Construction for Visible-Infrared Person Re-Identification.

Visible-infrared person re-identification (VI-ReID) is the task of matching the same individuals across the visible and infrared modalities. Its main challenge lies in the modality gap caused by the cameras operating on different spectra. Existing VI-ReID methods mainly focus on learning general features across modalities, often at the expense of feature discriminability. To address this issue, we present a novel cycle-construction-based network for neutral yet discriminative feature learning, termed CycleTrans. Specifically, CycleTrans uses a lightweight knowledge capturing module (KCM) to capture rich semantics from the modality-relevant feature maps according to pseudo anchors. Afterward, a discrepancy modeling module (DMM) is deployed to transform these features into neutral ones according to the modality-irrelevant prototypes. To ensure feature discriminability, another two KCMs are further deployed for feature cycle constructions. With cycle construction, our method can learn effective neutral features for visible and infrared images while preserving their salient semantics. Extensive experiments on SYSU-MM01 and RegDB datasets validate the merits of CycleTrans against a flurry of state-of-the-art (SOTA) methods, +1.88% on rank-1 in SYSU-MM01 and +1.1% on rank-1 in RegDB. Our code is available at https://github.com/DoubtedSteam/CycleTrans.

Clinical characteristics of autoimmune encephalitis with co-existence of multiple anti-neuronal antibodies.

BACKGROUND: An increasing number of cases of autoimmune encephalitis (AE) with co-existing multiple anti-neuronal antibodies have been reported in recent years. However, the clinical significance of the concurrent presence of multiple anti-neuronal antibodies in patients with AE remains unclear. METHODS: We retrospectively enrolled AE patients with multiple anti-neuronal antibodies treated at our center between August 2019 and February 2022. We also reviewed cases reported in multiple literature databases. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline was followed on selection process. And then the clinical and laboratory data of these cases were collected for review and summary. RESULTS: A total of 83 AE cases with multiple antibodies (9 cases from our center and 74 cases from the literatures reviewed) were identified. In our center, nine patients presented with encephalitis symptoms, clinically characterized as disturbed consciousness, seizures, cognitive impairment, and psychiatric disorders. Of the 83 cases, 73 cases had co-existence of 2 types of antibodies, 8 cases had 3 types, and 2 cases had 4 types. Thirty-nine cases (39/83, 46.9%) were confirmed or suspected of also having a tumor, of which the most common was lung cancer (28/83, 33.7%). Partial or complete recovery was achieved in 57 cases (57/83, 68.6%), while 26 cases (26/83, 31.3%) died during treatment or follow-up. CONCLUSIONS: AE with co-existing multiple anti-neuronal antibodies is a specific subgroup, that is increasingly recognized in clinical practice. The co-existence of multiple anti-neuronal antibodies has a major impact on clinical features, disease progression, and prognosis.

Field evaluation of a simple and rapid diagnostic test, RLDT to detect Shigella and enterotoxigenic E. coli in Indian children.

The diagnostic assays currently used to detect Shigella spp. (Shigella) and enterotoxigenic Escherichia coli (ETEC) are complex or elaborate which make them difficult to apply in resource poor settings where these diseases are endemic. The simple and rapid nucleic acid amplification-based assay "Rapid LAMP-based Diagnostic Test (RLDT)" was evaluated to detect Shigella spp (Shigella) and enterotoxigenic Escherichia coli (ETEC) and determine the epidemiology of these pathogens in Kolkata, India. Stool samples (n = 405) from children under five years old with diarrhea seeking care at the hospitals were tested, and 85(21%) and 68(17%) by RLDT, 91(23%) and 58(14%) by quantitative PCR (qPCR) and 35(9%) and 15(4%) by culture, were positive for Shigella and ETEC, respectively. The RLDT showed almost perfect agreement with qPCR, Kappa 0.96 and 0.89; sensitivity 93% and 98%; specificity 100% and 97% for Shigella and ETEC, respectively. While RLDT detected 12% more Shigella and 13% more ETEC than culture, all culture positives for Shigella and ETEC except one each were also positive by the RLDT, sensitivity 97% and 93% respectively. RLDT is a simple, sensitive, and rapid assay that could be implemented with minimum training in the endemic regions to strengthen the disease surveillance system and rapid outbreak detection.

Structural and mechanistic insights into the MCM8/9 helicase complex.

MCM8 and MCM9 form a functional helicase complex (MCM8/9) that plays an essential role in DNA homologous recombination repair for DNA double-strand break. However, the structural characterization of MCM8/9 for DNA binding/unwinding remains unclear. Here, we report structures of the MCM8/9 complex using cryo-electron microscopy single particle analysis. The structures reveal that MCM8/9 is arranged into a heterohexamer through a threefold symmetry axis, creating a central channel that accommodates DNA. Multiple characteristic hairpins from the N-terminal oligosaccharide/oligonucleotide (OB) domains of MCM8/9 protrude into the central channel and serve to unwind the duplex DNA. When activated by HROB, the structure of MCM8/9's N-tier ring converts its symmetry from C3 to C1 with a conformational change that expands the MCM8/9's trimer interface. Moreover, our structural dynamic analyses revealed that the flexible C-tier ring exhibited rotary motions relative to the N-tier ring, which is required for the unwinding ability of MCM8/9. In summary, our structural and biochemistry study provides a basis for understanding the DNA unwinding mechanism of MCM8/9 helicase in homologous recombination.

c-Met is a chimeric antigen receptor T-cell target for treating recurrent nasopharyngeal carcinoma.

BACKGROUND AIMS: Radiation therapy is the standard treatment for patients with nasopharyngeal carcinoma (NPC), but relapse occurs in 10% to 20% of patients. The treatment of recurrent nasopharyngeal carcinoma (rNPC) remains challenging. Chimeric antigen receptors (CAR)-T-cell therapy has achieved good outcomes in the treatment of leukemia and seems to be a promising therapeutic strategy for solid tumors. c-Met has been found to be highly expressed in multiple cancer types, and the activation of c-Met leads to the proliferation and metastasis of cancer cells. However, the expression of c-Met in rNPC tissues and whether it can be used as a target for CAR-T therapy in rNPC remain to be investigated. METHODS: We detected the expression of c-Met in 24 primary human rNPC tissues and three NPC cell lines and constructed two different antibody-derived anti-c-Met CARs, namely, Ab928z and Ab1028z. To estimate the function of these two different c-Met-targeted CAR-T cells, CD69 expression, cytotoxicity and cytokine secretion of CAR-T cells were assessed after coculture with target cells. A cell line-derived xenograft mouse model also was used to evaluate these two anti-c-Met CAR-T cells. Furthermore, we determined whether combination with an anti-EGFR antibody could promote the antitumor effect of CAR-T cells in a patient-derived xenograft mouse model. RESULTS: High c-Met expression was detected in 23 of 24 primary human rNPC tissues by immunohistochemistry staining and in three NPC cell lines by flow cytometry. Ab928z-T cells and Ab1028z-T cells showed significantly upregulated expression of CD69 after coculture with targeted cells. However, Ab1028z-T cells showed superior cytokine secretion and antitumor activity. Furthermore, Ab1028z-T cells effectively suppressed tumor growth compared with control CAR-T cells, and the combination with nimotuzumab further enhanced the tumor-clearing ability of Ab1028z-T cells. CONCLUSIONS: We found that c-Met is highly expressed in rNPC tissues and confirmed its potential as a CAR-T target for rNPC. Our study provides a new idea for the clinical treatment of rNPC.

Modeling the Interaction and Uptake of Cd-As(V) Mixture to Wheat Roots Affected by Humic Acids, in Terms of root cell Membrane Surface Potential (ψ0).

The joint toxicological effects of Cd2+ and As(V) mixture on wheat root as affected by environmental factors, such as pH, coexisting cations, and humic acids etc., were investigated using hydroponic experiments. The interaction and toxicological mechanisms of co-existing Cd2+ and As(V) at the interface of solution and roots in presence of humic acid were further explored by incorporating root cell membrane surface potential ψ0 into a mechanistic model of combined biotic ligand model (BLM)-based Gouy-Chapman-Stern (GCS) model and NICA-DONNAN model. Besides, molecular dynamics (MD) simulations of lipid bilayer equilibrated with solution containing Cd2+ and H2AsO4- further revealed the molecular distribution of heavy metal(loid) ions under different membrane surface potentials. H2AsO4- and Cd2+ can be adsorbed on the surface of the membrane alone or as complexes, which consolidate the limitation of the macroscopic physical models.

Anoikis-related genes combined with single cell sequencing: Insights into model specification of lung adenocarcinoma and applicability for prognosis and therapy.

Background: Anoikis has therapeutic potential against different malignancies including lung adenocarcinoma. This study used anoikis and bioinformatics to construct a prognostic model for lung adenocarcinoma and explore new therapeutic strategies. Methods: Several bioinformatic algorithms (co-expression analysis, univariate Cox analysis, multivariate Cox analysis, and cross-validation) were used to screen anoikis-related genes (ARGs) to construct a risk model. Lung adenocarcinoma patients were divided into training and testing groups at a ratio of 1:1. The prognostic model was validated by risk score comparison between high- and low-risk groups using receiver operating characteristic curve (ROC), nomograms, independent prognostic analysis and principal component analysis. In addition, two anoikis-related genes patterns were classified utilizing consensus clustering method and were compared with each other in survival time, immune microenvironment, and regulation in pathway. Single cell sequencing was applied to analyze anoikis-related genes constructed the model. Results: This study demonstrated the feasibility of the model based on seven anoikis-related genes, as well as identifying axitinib, nibtinib and sorafenib as potential therapeutic strategies for LUAD. Risk score based on this model had could be used as an independent prognostic factor for lung adenocarcinoma (HR > 1; p < 0.001) and had the highest accuracy to predict survival compared with the clinical characteristics. Single cell sequencing analysis discovered Keratin 14 (KRT14, one of the seven anoikis-related genes) was mainly expressed in malignant cells in various cancers. Conclusion: We identified seven anoikis-related genes and constructed an accurate risk model based on bioinformatics analysis that can be used for prognostic prediction and for the design of therapeutic strategies in clinical practice.

5-methyladenosine regulators play a crucial role in development of chronic hypersensitivity pneumonitis and idiopathic pulmonary fibrosis.

5-methyladenosine (m5C) modification regulates gene expression and biological functions in oncologic areas. However, the effect of m5C modification in chronic hypersensitivity pneumonitis (CHP) and idiopathic pulmonary fibrosis (IPF) remains unknown. Expression data for 12 significant m5C regulators were obtained from the interstitial lung disease dataset. Five candidate m5C regulators, namely tet methylcytosine dioxygenase 2, NOP2/Sun RNA methyltransferase 5, Y-box binding protein 1, tRNA aspartic acid methyltransferase 1, and NOP2/Sun RNA methyltransferase 3 were screened using random forest and nomogram models to predict risks of pulmonary fibrosis. Next, we applied the consensus clustering method to stratify the samples with different m5C patterns into two groups (cluster A and B). Finally, we calculated immune cell infiltration scores via single-sample gene set enrichment analysis, then compared immune cell infiltration, related functions as well as the expression of programmed cell death 1 (PD-1, PDCD1) and programmed death protein ligand-1 (PD-L1, CD274) between the two clusters. Principal component analysis of m5C-related scores across the 288 samples revealed that cluster A had higher immune-related expression than B. Notably, T helper cell (Th) 2 type cytokines and Th1 signatures were more abundant in clusters A and B, respectively. Our results suggest that m5C is associated with and plays a crucial role in development of pulmonary fibrosis. These m5C patterns could be potential biomarkers for identification of CHP and IPF, and guide future development of immunotherapy or other new drugs strategies for pulmonary fibrosis.

Diagnosis and Prognosis of Non-small Cell Lung Cancer based on Machine Learning Algorithms.

BACKGROUND: Non-small cell lung cancer (NSCLC) has been the subject of intense scholarly debate. We aimed to identify the potential biomarkers via bioinformatics analysis. METHODS: Three datasets were downloaded from gene expression omnibus database (GEO). R software was applied to screen differentially expressed genes (DEGs)and analyze immune cell infiltrates. Gene set enrichment analysis (GSEA) showed significant function and pathway in two groups. The diagnostic markers were further investigated by multiple machine learning algorithms (least absolute shrinkage and selection operator (LASSO) and support vector machine-recursive feature elimination (SVM-RFE)). Various online analytic platforms were utilized to explore the expression and prognostic value of differential genes. Furthermore, western blotting was performed to test the effects of genes on cell proliferation in vitro. RESULTS: We identified 181 DEGs shared by two datasets and selected nine diagnostic markers. Those genes were also significantly overexpressed in the third dataset. Topoisomerase II alpha (TOP2A) is overexpressed in lung cancer and associated with a poor prognosis, which was confirmed using immunohistochemistry (IHC) and Western blotting. Additionally, TOP2A showed a negative correlation with immune cells, such as CD8+ T cells, eosinophils and natural killer (NK) cell. CONCLUSION: Collectively, for the first time, we applied multiple machine learning algorithms, online databases and experiments in vitro to show that TOP2A is a potential biomarker for lung adenocarcinoma and could facilitate the development of new treatment strategies.

A Real-Time Global Inference Network for One-Stage Referring Expression Comprehension.

Referring expression comprehension (REC) is an emerging research topic in computer vision, which refers to the detection of a target region in an image given a test description. Most existing REC methods follow a multistage pipeline, which is computationally expensive and greatly limits the applications of REC. In this article, we propose a one-stage model toward real-time REC, termed real-time global inference network (RealGIN). RealGIN addresses the issues of expression diversity and complexity of REC with two innovative designs: adaptive feature selection (AFS) and Global Attentive ReAsoNing (GARAN). Expression diversity concerns varying expression content, which includes information such as colors, attributes, locations, and fine-grained categories. To address this issue, AFS adaptively fuses features of different semantic levels to tackle the changes in expression content. In contrast, expression complexity concerns the complex relational conditions in expressions that are used to identify the referent. To this end, GARAN uses the textual feature as a pivot to collect expression-aware visual information from all regions and then diffuses this information back to each region, which provides sufficient context for modeling the relational conditions in expressions. On five benchmark datasets, i.e., RefCOCO, RefCOCO+, RefCOCOg, ReferIT, and Flickr30k, the proposed RealGIN outperforms most existing methods and achieves very competitive performances against the most advanced one, i.e., MAttNet. More importantly, under the same hardware, RealGIN can boost the processing speed by 10-20 times over the existing methods.

Corrigendum: Relevance of RNA N6-methyladenosine regulators for pulmonary fibrosis: Implications for chronic hypersensitivity pneumonitis and idiopathic pulmonary fibrosis.

[This corrects the article DOI: 10.3389/fgene.2022.939175.].

An Analysis of BMP1 Associated with m6A Modification and Immune Infiltration in Pancancer.

Background: This research explores the underlying link between diagnosis and therapy between bone morphogenetic protein 1 (BMP1) and various cancers. Methods: Three immunotherapeutic cohorts, by the composition of IMvigor210, GSE35640, and GSE78220 were obtained from previously published articles and the Gene Expression Omnibus database. The different expressions of BMP1 in various clinical parameters were conducted, and prognostic analysis was executed utilizing Cox proportional hazard regression and Gene Expression Profiling Interactive Analysis. Moreover, the correlation between BMP1 and tumor microenvironment was analyzed using ESTIMATE and CIBERSORT algorithms. Tumor mutational burden and microsatellite instability were also included. The correlation between m6A modification and the gene expression level was analyzed using Tumor IMmune Estimation Resource, the University of Alabama at Birmingham Cancer data analysis portal. Gene Set Cancer Analysis analyzed the correlation of BMP1 expression level with copy number variations and methylation. Furthermore, the correlation between BMP1 and therapeutic response after antineoplastic drug use was illustrated for further discussion. Results: BMP1 expression had significant differences in 14 cancers. It presented an intimate relationship with immune-relevant biomarkers. A variation analysis indicated that BMP1 had a significant association with immunotherapeutic response. The expression level of BMP1 was closely associated with insulin-like growth factor binding protein 3, an m6A modification relative gene. Except for a few cancer types, methylation negatively correlated with BMP1, and copy number variations positively correlated with BMP1. Notably, low BMP1 expression was connected with immunotherapeutic response in the cohorts, and its expression was related to increased sectional sensitivity of drugs. Conclusion: BMP1 may serve as a potential biomarker for prognostic prediction and immunologic infiltration in diversified cancers, providing a new thought approach for oncotherapy.

Urine cytological study in patients with clinicopathologically confirmed neuronal intranuclear inclusion disease.

Objective: The diagnosis of neuronal intranuclear inclusion disease (NIID) is currently based on CGG repeat expansion in the 5'UTR of the NOTCH2NLC gene, or p62-positive intranuclear inclusions in skin biopsy. The purpose of this study is to explore the value of non-invasive pathological findings in urine sediment cells from NIID patients. Materials and methods: Ten patients with clinically suspected NIID were enrolled for skin biopsy and gene screening. Morning urine (500 ml) was collected from each patient, and cell sediment was obtained by centrifugation. Urine cytology, including Giemsa staining, p62 immunostaining, and electron microscopic examination, were conducted on cell sediment. Results: The main clinical symptoms of 10 patients included episodic disturbance of consciousness, cognitive impairment, tremor, limb weakness, and so on. Cerebral MRI showed that 9 patients had linear DWI high signal in the corticomedullary junction. Genetic testing found that the number of CGG repeat ranged from 96 to 158 in the NOTCH2NLC gene. Skin biopsy revealed that all patients showed p62-positive intranuclear inclusions in 18.5 ± 6.3% of the duct epithelial cells of sweat gland. In contrast, urine sediment smears revealed that only 3 patients had p62 positive intranuclear inclusions in 3.5 ± 1.2% of the sedimentary cells. Ultrastructural examinations showed that intranuclear inclusions were also identified in the cell sediment of the 3 patients. Conclusion: Urine cytology may be a new and non-invasive pathological diagnosis technique for some NIID patients, although the positive rate is not as high as that of skin biopsy, which is a sensitive and reliable pathological method for NIID.

Relevance of RNA N6-Methyladenosine Regulators for Pulmonary Fibrosis: Implications for Chronic Hypersensitivity Pneumonitis and Idiopathic Pulmonary Fibrosis.

N6-methyladenosine (m6A) modification plays a pivotal role in post-transcriptionally regulating gene expression and biological functions. Nonetheless, the roles of m6A modification in the regulation of chronic hypersensitivity pneumonitis (CHP) and idiopathic pulmonary fibrosis (IPF) remain unclear. Twenty-two significant m6A regulators were selected from differential gene analysis between the control and treatment groups from the GSE150910 dataset. Five candidate m6A regulators (insulin-like growth factor binding protein 2, insulin-like growth factor binding protein 3, YTH domain-containing protein 1, zinc finger CCCH domain-containing protein 13, and methyltransferase-like 3) were screened by the application of a random forest model and nomogram model to predict risks of pulmonary fibrosis. The consensus clustering method was applied to divide the treatment samples into two groups with different m6A patterns (clusters A and B) based on the 22 m6A regulators. Our study performed principal component analysis to obtain the m6A-related score of the 288 samples to quantify the two m6A patterns. The study reveals that cluster A was linked to T helper cell (Th) 2-type cytokines, while the immune infiltration of Th1 cytokines was higher in cluster B. Our results suggest that m6A cluster A is likely related to pulmonary fibrosis, indicating m6A regulators play notable roles in the occurrence of pulmonary fibrosis. The m6A patterns could be considered as biomarkers to identify CHP and IPF, which will be helpful to develop immunotherapy strategies for pulmonary fibrosis in the future.

Knowing What to Learn: A Metric-Oriented Focal Mechanism for Image Captioning.

Despite considerable progress, image captioning still suffers from the huge difference in quality between easy and hard examples, which is left unexploited in existing methods. To address this issue, we explore the hard example mining in image captioning, and propose a simple yet effective mechanism to instruct the model to pay more attention to hard examples, thereby improving the performance in both general and complex scenarios. We first propose a novel learning strategy, termed Metric-oriented Focal Mechanism (MFM), for hard example mining in image captioning. Differing from the existing strategies for classification tasks, MFM can adopt the generative metrics of image captioning to measure the difficulties of examples, and then up-weight the rewards of hard examples during training. To make MFM applicable to different datasets without tedious parameter tuning, we further introduce an adaptive reward metric called Effective CIDEr (ECIDEr), which considers the data distribution of easy and hard examples during reward estimation. Extensive experiments are conducted on the MS COCO benchmark, and the results show that while maintaining the performance on simple examples, MFM can significantly improve the quality of captions for hard examples. The ECIDEr-based MFM is equipped on the current SOTA method, e.g., DLCT (Luo et al., 2021), which outperforms all existing methods and achieves new state-of-the-art performance on both the off-line and on- line testing, i.e., 134.3 CIDEr for the off-line testing and 136.1 for the on- line testing of MSCOCO. To validate the generalization ability of ECIDEr-based MFM, we also apply it to another dataset, namely Flickr30k, and superior performance gains can also be obtained.

Towards Lightweight Transformer Via Group-Wise Transformation for Vision-and-Language Tasks.

Despite the exciting performance, Transformer is criticized for its excessive parameters and computation cost. However, compressing Transformer remains as an open problem due to its internal complexity of the layer designs, i.e., Multi-Head Attention (MHA) and Feed-Forward Network (FFN). To address this issue, we introduce Group-wise Transformation towards a universal yet lightweight Transformer for vision-and-language tasks, termed as LW-Transformer. LW-Transformer applies Group-wise Transformation to reduce both the parameters and computations of Transformer, while also preserving its two main properties, i.e., the efficient attention modeling on diverse subspaces of MHA, and the expanding-scaling feature transformation of FFN. We apply LW-Transformer to a set of Transformer-based networks, and quantitatively measure them on three vision-and-language tasks and six benchmark datasets. Experimental results show that while saving a large number of parameters and computations, LW-Transformer achieves very competitive performance against the original Transformer networks for vision-and-language tasks. To examine the generalization ability, we apply LW-Transformer to the task of image classification, and build its network based on a recently proposed image Transformer called Swin-Transformer, where the effectiveness can be also confirmed.

Circulating NAD+ Metabolism-Derived Genes Unveils Prognostic and Peripheral Immune Infiltration in Amyotrophic Lateral Sclerosis.

Background: Nicotinamide adenine dinucleotide (NAD+) metabolism has drawn more attention on neurodegeneration research; however, the role in Amyotrophic Lateral Sclerosis (ALS) remains to be fully elucidated. Here, the purpose of this study was to investigate whether the circulating NAD+ metabolic-related gene signature could be identified as a reliable biomarker for ALS survival. Methods: A retrospective analysis of whole blood transcriptional profiles and clinical characteristics of 454 ALS patients was conducted in this study. A series of bioinformatics and machine-learning methods were combined to establish NAD+ metabolic-derived risk score (NPRS) to predict overall survival for ALS patients. The associations of clinical characteristic with NPRS were analyzed and compared. Receiver operating characteristic (ROC) and the calibration curve were utilized to assess the efficacy of prognostic model. Besides, the peripheral immune cell infiltration was assessed in different risk subgroups by applying the CIBERSORT algorithm. Results: Abnormal activation of the NAD+ metabolic pathway occurs in the peripheral blood of ALS patients. Four subtypes with distinct prognosis were constructed based on NAD+ metabolism-related gene expression patterns by using the consensus clustering method. A comparison of the expression profiles of genes related to NAD+ metabolism in different subtypes revealed that the synthase of NAD+ was closely associated with prognosis. Seventeen genes were selected to construct prognostic risk signature by LASSO regression. The NPRS exhibited stronger prognostic capacity compared to traditional clinic-pathological parameters. High NPRS was characterized by NAD+ metabolic exuberant with an unfavorable prognosis. The infiltration levels of several immune cells, such as CD4 naive T cells, CD8 T cells, neutrophils and macrophages, are significantly associated with NPRS. Further clinicopathological analysis revealed that NPRS is more appropriate for predicting the prognostic risk of patients with spinal onset. A prognostic nomogram exhibited more accurate survival prediction compared with other clinicopathological features. Conclusions: In conclusion, it was first proposed that the circulating NAD+ metabolism-derived gene signature is a promising biomarker to predict clinical outcomes, and ultimately facilitating the precise management of patients with ALS.