Constructing lactylation-related genes prognostic model to effectively predict the disease-free survival and treatment responsiveness in prostate cancer based on machine learning.

Background: Prostate cancer (PCa) is one of the most common malignancies in men with a poor prognosis. It is therefore of great clinical importance to find reliable prognostic indicators for PCa. Many studies have revealed the pivotal role of protein lactylation in tumor development and progression. This research aims to analyze the effect of lactylation-related genes on PCa prognosis. Methods: By downloading mRNA-Seq data of TCGA PCa, we obtained the differential genes related to lactylation in PCa. Five machine learning algorithms were used to screen for lactylation-related key genes for PCa, then the five overlapping key genes were used to construct a survival prognostic model by lasso cox regression analysis. Furthermore, the relationships between the model and related pathways, tumor mutation and immune cell subpopulations, and drug sensitivity were explored. Moreover, two risk groups were established according to the risk score calculated by the five lactylation-related genes (LRGs). Subsequently, a nomogram scoring system was established to predict disease-free survival (DFS) of patients by combining clinicopathological features and lactylation-related risk scores. In addition, the mRNA expression levels of five genes were verified in PCa cell lines by qPCR. Results: We identified 5 key LRGs (ALDOA, DDX39A, H2AX, KIF2C, RACGAP1) and constructed the LRGs prognostic model. The AUC values for 1 -, 3 -, and 5-year DFS in the TCGA dataset were 0.762, 0.745, and 0.709, respectively. The risk score was found a better predictor of DFS than traditional clinicopathological features in PCa. A nomogram that combined the risk score with clinical variables accurately predicted the outcome of the patients. The PCa patients in the high-risk group have a higher proportion of regulatory T cells and M2 macrophage, a higher tumor mutation burden, and a worse prognosis than those in the low-risk group. The high-risk group had a lower IC50 for certain chemotherapeutic drugs, such as Docetaxel, and Paclitaxel than the low-risk group. Furthermore, five key LRGs were found to be highly expressed in castration-resistant PCa cells. Conclusion: The lactylation-related genes prognostic model can effectively predict the DFS and therapeutic responses in patients with PCa.

Deep learning to predict lymph node status on pre-operative staging CT in patients with colon cancer.

INTRODUCTION: Lymph node (LN) metastases are an important determinant of survival in patients with colon cancer, but remain difficult to accurately diagnose on preoperative imaging. This study aimed to develop and evaluate a deep learning model to predict LN status on preoperative staging CT. METHODS: In this ambispective diagnostic study, a deep learning model using a ResNet-50 framework was developed to predict LN status based on preoperative staging CT. Patients with a preoperative staging abdominopelvic CT who underwent surgical resection for colon cancer were enrolled. Data were retrospectively collected from February 2007 to October 2019 and randomly separated into training, validation, and testing cohort 1. To prospectively test the deep learning model, data for testing cohort 2 was collected from October 2019 to July 2021. Diagnostic performance measures were assessed by the AUROC. RESULTS: A total of 1,201 patients (median [range] age, 72 [28-98 years]; 653 [54.4%] male) fulfilled the eligibility criteria and were included in the training (n = 401), validation (n = 100), testing cohort 1 (n = 500) and testing cohort 2 (n = 200). The deep learning model achieved an AUROC of 0.619 (95% CI 0.507-0.731) in the validation cohort. In testing cohort 1 and testing cohort 2, the AUROC was 0.542 (95% CI 0.489-0.595) and 0.486 (95% CI 0.403-0.568), respectively. CONCLUSION: A deep learning model based on a ResNet-50 framework does not predict LN status on preoperative staging CT in patients with colon cancer.

A new secondary metabolite from the marine-derived fungus Phomopsis lithocarpus FS508.

Based on the One Strain-Many Compounds (OSMAC) strategy, the secondary metabolites of Phomopsis lithocarpus FS508 were investigated. As a result, a new secondary metabolite, 4-methoxy-3-[4-(acetyloxy)-3-methyl-2-butenyl]benzoic acid (1) as well as eleven known compounds were isolated from the fermentation product of the strain FS508. Their structures were determined by NMR, IR, UV, and MS spectroscopic data analyses. All the isolated compounds were evaluated for cytotoxic and anti-inflammatory activities. Among them, compounds 3 and 9 displayed potent cytotoxicity against HepG-2 cell line, and compounds 2, 3 and 12 showed significant anti-inflammatory activities.

Pimarane-Type Diterpenes with Anti-Inflammatory Activity from Arctic-Derived Fungus Eutypella sp. D-1.

The Arctic-derived fungus Eutypella sp. D-1 can produce numerous secondary metabolites, and some compounds exhibit excellent biological activity. Seven pimarane-type diterpenes, including three new compounds eutypellenone F (1), libertellenone Y (2), and libertellenone Z (3), and four known compounds (4-7), were isolated from fermentation broth of Eutypella sp. D-1 by the OSMAC strategy of adding ethanol as a promoter in the culture medium. Compound 2 has a rare tetrahydrofuran-fused pimarane diterpene skeleton. The anti-inflammatory activity of all compounds was evaluated. Compounds 3-6 showed a significant inhibitory effect on cell NO release at 10 µmol/L by in vitro experiments, of which 3-5 had inhibitory rates over 60% on nitric oxide (NO) release. Subsequently, the anti-inflammatory activity of 3-5 was evaluated based on a zebrafish model, and the results showed that 3 had a significant inhibitory effect on inflammatory cells migration at 40 µmol/L, while 4 and 5 had a significant inhibitory effect at 20 µmol/L. Moreover, compounds 3-5 have the same conjugated double bond structure, which may be an important group for these compounds to exert anti-inflammatory activity.

Learning From Partially Labeled Data for Multi-Organ and Tumor Segmentation.

Medical image benchmarks for the segmentation of organs and tumors suffer from the partially labeling issue due to its intensive cost of labor and expertise. Current mainstream approaches follow the practice of one network solving one task. With this pipeline, not only the performance is limited by the typically small dataset of a single task, but also the computation cost linearly increases with the number of tasks. To address this, we propose a Transformer based dynamic on-demand network (TransDoDNet) that learns to segment organs and tumors on multiple partially labeled datasets. Specifically, TransDoDNet has a hybrid backbone that is composed of the convolutional neural network and Transformer. A dynamic head enables the network to accomplish multiple segmentation tasks flexibly. Unlike existing approaches that fix kernels after training, the kernels in the dynamic head are generated adaptively by the Transformer, which employs the self-attention mechanism to model long-range organ-wise dependencies and decodes the organ embedding that can represent each organ. We create a large-scale partially labeled Multi-Organ and Tumor Segmentation benchmark, termed MOTS, and demonstrate the superior performance of our TransDoDNet over other competitors on seven organ and tumor segmentation tasks. This study also provides a general 3D medical image segmentation model, which has been pre-trained on the large-scale MOTS benchmark and has demonstrated advanced performance over current predominant self-supervised learning methods.

A New Signature That Predicts Progression-Free Survival of Clear Cell Renal Cell Carcinoma with Anti-PD-1 Therapy.

Immunotherapy has greatly improved the survival time and quality of life of patients with renal cell carcinoma, but the benefits are limited to a small portion of patients. There are too few new biomarkers that can be used to identify molecular subtypes of renal clear cell carcinoma and predict survival time with anti-PD-1 treatment. Single-cell RNA data of clear cell renal cell carcinoma (ccRCC) treated with anti-PD-1 were obtained from public databases, then 27,707 high-quality CD4 + T and CD8 + T cells were obtained for subsequent analysis. Firstly, genes set variation analysis and CellChat algorithm were used to explore potential molecular pathway differences and intercellular communication between the responder and non-responder groups. Additionally, differentially expressed genes (DEGs) between the responder and non-responder groups were obtained using the "edgeR" package, and ccRCC samples from TCGA-KIRC (n = 533) and ICGA-KIRC (n = 91) were analyzed by the unsupervised clustering algorithm to recognize molecular subtypes with different immune characteristics. Finally, using univariate Cox analysis, least absolute shrinkage and selection operator (Lasso) regression, and multivariate Cox regression, the prognosis model of immunotherapy was established and verified to predict the progression-free survival of ccRCC patients treated with anti-PD-1. At the single cell level, there are different signal pathways and cell communication between the immunotherapy responder and non-responder groups. In addition, our research also confirms that the expression level of PDCD1/PD-1 is not an effective marker for predicting the response to immune checkpoint inhibitors (ICIs). The new prognostic immune signature (PIS) enabled the classification of ccRCC patients with anti-PD-1 therapy into high- and low-risk groups, and the progression-free survival times (PFS) and immunotherapy responses were significantly different between these two groups. In the training group, the area under the ROC curve (AUC) for predicting 1-, 2- and 3-year progression-free survival was 0.940 (95% CI: 0.894-0.985), 0.981 (95% CI: 0.960-1.000), and 0.969 (95% CI: 0.937-1.000), respectively. Validation sets confirm the robustness of the signature. This study revealed the heterogeneity between the anti-PD-1 responder and non-responder groups from different angles and established a robust PIS to predict the progression-free survival of ccRCC patients receiving immune checkpoint inhibitors.

The Natural Products and Extracts: Anti-Triple-Negative Breast Cancer in Vitro.

Triple-negative breast cancer (TNBC) makes up 15 % to 20 % of all breast cancer (BC) cases, and represents one of the most challenging malignancies to treat. For many years, chemotherapy has been the main treatment option for TNBC. Natural products isolated from marine organisms and terrestrial organisms with great structural diversity and high biochemical specificity form a compound library for the assessment and discovery of new drugs. In this review, we mainly focused on natural compounds and extracts (from marine and terrestrial environments) with strong anti-TNBC activities (IC50 <100 µM) and their possible mechanisms reported in the past six years (2015-2021).

Cytotoxic meroterpenoids from the marine sponge Dactylospongia elegans.

Three new sesquiterpene quinones/hydroquinones, 20-demethoxy-20-isopentylaminodactyloquinone D (1), 20-demethoxy-20-isobutylaminodactyloquinone D (2), and 19-methoxy-dictyoceratin-A (3), and five known related compounds (4-8) were isolated from the marine sponge Dactylospongia elegans. Their structures were elucidated by spectroscopic analysis, ECD calculation, single-crystal X-ray diffraction, and comparison with the literature. Compounds 3 and 5-8 exhibited activities against the human cancer cell lines DU145, SW1990, Huh7, and PANC-1 with IC50 values ranging from 2.33 to 37.85 µM.

Whole-genome sequencing to detect mutations associated with resistance to insecticides and Bt proteins in Spodoptera frugiperda.

The fall armyworm (FAW), Spodoptera frugiperda, is a major pest native to the Americas that has recently invaded the Old World. Point mutations in the target-site proteins acetylcholinesterase-1 (ace-1), voltage-gated sodium channel (VGSC) and ryanodine receptor (RyR) have been identified in S. frugiperda as major resistance mechanisms to organophosphate, pyrethroid and diamide insecticides respectively. Mutations in the adenosine triphosphate-binding cassette transporter C2 gene (ABCC2) have also been identified to confer resistance to Cry1F protein. In this study, we applied a whole-genome sequencing (WGS) approach to identify point mutations in the target-site genes in 150 FAW individuals collected from China, Malawi, Uganda and Brazil. This approach revealed three amino acid substitutions (A201S, G227A and F290V) of S. frugiperda ace-1, which are known to be associated with organophosphate resistance. The Brazilian population had all three ace-1 point mutations and the 227A allele (mean frequency = 0.54) was the most common. Populations from China, Malawi and Uganda harbored two of the three ace-1 point mutations (A201S and F290V) with the 290V allele (0.47-0.58) as the dominant allele. Point mutations in VGSC (T929I, L932F and L1014F) and RyR (I4790M and G4946E) were not detected in any of the 150 individuals. A novel 12-bp insertion mutation in exon 15 of the ABCC2 gene was identified in some of the Brazilian individuals but absent in the invasive populations. Our results not only demonstrate robustness of the WGS-based genomic approach for detection of resistance mutations, but also provide insights for improvement of resistance management tactics in S. frugiperda.

Inter-Slice Context Residual Learning for 3D Medical Image Segmentation.

Automated and accurate 3D medical image segmentation plays an essential role in assisting medical professionals to evaluate disease progresses and make fast therapeutic schedules. Although deep convolutional neural networks (DCNNs) have widely applied to this task, the accuracy of these models still need to be further improved mainly due to their limited ability to 3D context perception. In this paper, we propose the 3D context residual network (ConResNet) for the accurate segmentation of 3D medical images. This model consists of an encoder, a segmentation decoder, and a context residual decoder. We design the context residual module and use it to bridge both decoders at each scale. Each context residual module contains both context residual mapping and context attention mapping, the formal aims to explicitly learn the inter-slice context information and the latter uses such context as a kind of attention to boost the segmentation accuracy. We evaluated this model on the MICCAI 2018 Brain Tumor Segmentation (BraTS) dataset and NIH Pancreas Segmentation (Pancreas-CT) dataset. Our results not only demonstrate the effectiveness of the proposed 3D context residual learning scheme but also indicate that the proposed ConResNet is more accurate than six top-ranking methods in brain tumor segmentation and seven top-ranking methods in pancreas segmentation.

Depletion of SNRNP200 inhibits the osteo-/dentinogenic differentiation and cell proliferation potential of stem cells from the apical papilla.

BACKGROUND: Tissue regeneration mediated by mesenchymal stem cells (MSCs) is deemed a desirable way to repair teeth and craniomaxillofacial tissue defects. Nevertheless, the molecular mechanisms about cell proliferation and committed differentiation of MSCs remain obscure. Previous researches have proved that lysine demethylase 2A (KDM2A) performed significant function in the regulation of MSC proliferation and differentiation. SNRNP200, as a co-binding factor of KDM2A, its potential effect in regulating MSCs' function is still unclear. Therefore, stem cells from the apical papilla (SCAPs) were used to investigate the function of SNRNP200 in this research. METHODS: The alkaline phosphatase (ALP) activity assay, Alizarin Red staining, and osteogenesis-related gene expressions were used to examine osteo-/dentinogenic differentiation potential. Carboxyfluorescein diacetate, succinimidyl ester (CFSE) and cell cycle analysis were applied to detect the cell proliferation. Western blot analysis was used to evaluate the expressions of cell cycle-related proteins. RESULTS: Depletion of SNRNP200 caused an obvious decrease of ALP activity, mineralization formation and the expressions of osteo-/dentinogenic genes including RUNX2, DSPP, DMP1 and BSP. Meanwhile, CFSE and cell cycle assays revealed that knock-down of SNRNP200 inhibited the cell proliferation and blocked cell cycle at the G2/M and S phase in SCAPs. In addition, it was found that depletion of SNRNP200 up-regulated p21 and p53, and down-regulated the CDK1, CyclinB, CyclinE and CDK2. CONCLUSIONS: Depletion of SNRNP200 repressed osteo-/dentinogenic differentiation potentials and restrained cell proliferation through blocking cell cycle progression at the G2/M and S phase, further revealing that SNRNP200 has crucial effects on preserving the proliferation and differentiation potentials of dental tissue-derived MSCs.

Knockdown EIF3C Suppresses Cell Proliferation and Increases Apoptosis in Pancreatic Cancer Cell.

Increasing evidence shows that eukaryotic initiation factor subunit (EIF3C) plays a crucial role in development of tumors. However, the underlying roles of EIF3Cin the development of pancreatic cancer (PC) remain unknown. In this study, we examined the expression of EIF3C in PC tissues, their adjacent normal tissues and 3 cell lines (SW1990, PANC-1 and AsPC-1). Moreover, the EIF3C-shRNA lentivirus was constructed to suppress EIF3C expression. Following this, the cell colony formation assay was employed to evaluate proliferation ability of PC cells. Meanwhile, the cell cycle and apoptotic assays were also performed by flow cytometry. We found that level of EIF3C in PC tissues was significantly increased compared with that in adjacent normal tissues. Furthermore, the knockdown of EIF3C can significantly reduce cell proliferation, block cell cycle in G2/M and induce apoptosis in both SW1990 and PANC-1 cells. Our findings suggest that EIF3C plays a crucial role in the progression of PC and may be a potential target in the treatment of PC.

DLX5 and HOXC8 enhance the chondrogenic differentiation potential of stem cells from apical papilla via LINC01013.

BACKGROUND: Mesenchymal stem cell (MSC)-based cartilage tissue regeneration is a treatment with great potential. How to enhance the MSC chondrogenic differentiation is a key issue involved in cartilage formation. In the present study, we seek to expound the phenotypes and mechanisms of DLX5 in chondrogenic differentiation function in MSCs. METHODS: Stem cells from apical papilla (SCAPs) were used. The Alcian Blue staining, pellet culture system, and cell transplantation in rabbit knee cartilage defect were used to evaluate the chondrogenic differentiation function of MSCs. Western blot, real-time RT-PCR, and ChIP assays were used to evaluate the molecular mechanisms. RESULTS: DLX5 and HOXC8 expressions were upregulated during chondrogenic differentiation. In vitro results showed that DLX5 and HOXC8 enhanced the expression of chondrogenic markers including collagen II (COL2), collagen V (COL5), and sex-determining region Y box protein 9 (SOX9) and promoted the chondrogenic differentiation and the formation of cartilage clumps in the pellet culture system. Mechanically, DLX5 and HOXC8 formed protein complexes and negatively regulated the LncRNA, LINC01013, via directly binding its promoter. In vivo transplantation experiment showed that DLX5 and HOXC8 could restore the cartilage defect in the rabbit knee model. In addition, knock-down of LINC01013 enhanced the chondrogenic differentiation of SCAPs. CONCLUSIONS: In conclusion, DLX5 and HOXC8 enhance the chondrogenic differentiation abilities of SCAPs by negatively regulating LINC01013 in SCAPs, and provided the potential target for promoting cartilage tissue regeneration.

THBS2, a microRNA-744-5p target, modulates MMP9 expression through CUX1 in pancreatic neuroendocrine tumors.

The underlying molecular mechanisms of pancreatic neuroendocrine tumor (pNET) development have not yet been clearly identified. The present study revealed that thrombospondin 2 (THBS2) was downregulated in pNET tissues and cells. Forced expression of THBS2 inhibited the proliferation and migration of pNET cells in vitro. MicroRNA(miR)-744-5p was indicated to be a direct regulator of THBS2. Upregulation of miR-744-5p potentially caused THBS2 repression. Furthermore, THBS2 inhibited the production of matrix metalloproteinase (MMP) MMP9 through suppressing the transcriptional activity of CUT-like homeobox 1 (CUX1). CUX1 and MMP9 mediated the effect of THBS2 on pNET proliferation and migration, respectively. The results of the present study revealed a mechanistic role for THBS2 in pNET proliferation and migration, indicating that THBS2 was downregulated by miR-744-5p and further affected the CUX1/MMP9 cascade, which promoted the development of pNET.

A Mutual Bootstrapping Model for Automated Skin Lesion Segmentation and Classification.

Automated skin lesion segmentation and classification are two most essential and related tasks in the computer-aided diagnosis of skin cancer. Despite their prevalence, deep learning models are usually designed for only one task, ignoring the potential benefits in jointly performing both tasks. In this paper, we propose the mutual bootstrapping deep convolutional neural networks (MB-DCNN) model for simultaneous skin lesion segmentation and classification. This model consists of a coarse segmentation network (coarse-SN), a mask-guided classification network (mask-CN), and an enhanced segmentation network (enhanced-SN). On one hand, the coarse-SN generates coarse lesion masks that provide a prior bootstrapping for mask-CN to help it locate and classify skin lesions accurately. On the other hand, the lesion localization maps produced by mask-CN are then fed into enhanced-SN, aiming to transfer the localization information learned by mask-CN to enhanced-SN for accurate lesion segmentation. In this way, both segmentation and classification networks mutually transfer knowledge between each other and facilitate each other in a bootstrapping way. Meanwhile, we also design a novel rank loss and jointly use it with the Dice loss in segmentation networks to address the issues caused by class imbalance and hard-easy pixel imbalance. We evaluate the proposed MB-DCNN model on the ISIC-2017 and PH2 datasets, and achieve a Jaccard index of 80.4% and 89.4% in skin lesion segmentation and an average AUC of 93.8% and 97.7% in skin lesion classification, which are superior to the performance of representative state-of-the-art skin lesion segmentation and classification methods. Our results suggest that it is possible to boost the performance of skin lesion segmentation and classification simultaneously via training a unified model to perform both tasks in a mutual bootstrapping way.

Epiregulin promotes osteogenic differentiation and inhibits neurogenic trans-differentiation of adipose-derived mesenchymal stem cells via MAPKs pathway.

Mesenchymal stem cells (MSCs) exists low efficiency to trans-differentiate into other germinal layer cell types. One key issue is to discover the effect of important factor on MSCs differentiation abiltiy. In this study, we investigated the role and mechanism of epiregulin (EREG) on the osteogenic differentiation and neurogenic trans-differentiation in adipose-derived stem cells (ADSCs). We discovered that the depletion of EREG inhibited the osteogenic differentiation in vitro. And 25 ng/mL recombinant human epiregulin protein (rhEREG) effectively improved the osteogenic differentiation of EREG-depleted-ADSCs. Depletion of EREG promoted the formation of neural spheres, and increased the expressions of nestin, ßIII-tubulin, NeuroD, NCAM, TH, and NEF in ADSCs. Then, 25 ng/mL rhEREG significantly inhibited these neurogenic differentiation indicators. Inhibition of p38 MAPK, JNK, or Erk1/2 signaling pathway separately, blocked the rhEREG-enhanced osteogenic differentiation ability and the rhEREG-inhibited neurogenic trans-differentiation ability of ADSCs. In conclusions, EREG promoted the osteogenic differentiation and inhibited the neurogenic trans-differentiation potentials of ADSCs via MAPK signaling pathways.

Surgical treatment ofpatients with severe non-flail chest rib fractures.

BACKGROUND: Many patients have inadequate long-term analgesia, respiratory distress, and hypoxemia due to a long-standing substantial smoking history or the presence of primary pulmonary diseases; analgesic treatment is not valid in these patients. Even if the imaging findings of rib fractures are relatively mild, rib fractures may cause severe position limitation, respiratory distress, and hypoxemia. AIM: To investigate the curative effect of surgical treatment for patients with severe non-flail chest rib fractures. METHODS: A total of 78 patients from our hospital with severe noncontinuous thoracic rib fractures from September 2016 to September 2018 were enrolled in our study. Thirty-nine patients underwent surgical treatment, and 39 underwent conservative treatment. The surgical treatment group received surgery performed with titanium plates, and the screws were inserted with open reduction and internal fixation. The conservative treatment group received analgesia and symptomatic treatment. The pain scores at 72 h, 1 wk, 2 wk, 4 wk, 6 wk, 3 mo, and 6 mo were compared, and the SF-36 quality of life scores were compared atthe 3rd and 6th months. RESULTS: Pain relief in the surgical group was significantly better than that in the conservative group at each time point (72 h, 1 wk, 2 wk, 4 wk, 6 wk, 3 mo, and 6 mo after surgery, P < 0.001). ( The SF-36 scores were significantly higher in the surgical group than in the conservative group at 1 mo and 6 mo (P < 0.05). CONCLUSION: Patients with severe non-flail chest rib fractures have a better quality of life following surgical treatment than following conservative treatment, and surgical treatment is also useful for relieving pain. We should pay more attention to the physiological functions and clinical manifestations of patients with severe rib fractures. In patients with non-flail chest rib fractures, surgical treatment is feasible and effective.

Eutypellacytosporins A-D, Meroterpenoids from the Arctic Fungus Eutypella sp. D-1.

The Arctic fungus Eutypella sp. D-1, previously found to produce a variety of cytotoxic cyclopropyl-fused and cyclobutyl-fused pimarane diterpenoids when grown in the defined medium, was induced to produce unusual metabolites by growing on solid rice medium. A chemical investigation on the rice medium extract led to the isolation of four new meroterpenoids, eutypellacytosporins A-D (1-4), along with the known biogenetically related compound cytosporin D (5). The structures of the new compounds were elucidated by their detailed spectroscopic analysis and modified Mosher's method. Compounds 1-4 may be formed by the 12,32-ester linkage of two moieties, cytosporin D (5) and decipienolide A or B. All isolated compounds, except 5, showed weak cytotoxicity against DU145, SW1990, Huh7, and PANC-1 cell lines with IC50 values ranging from 4.9 to 17.1 µM.

Semi-supervised adversarial model for benign-malignant lung nodule classification on chest CT.

Classification of benign-malignant lung nodules on chest CT is the most critical step in the early detection of lung cancer and prolongation of patient survival. Despite their success in image classification, deep convolutional neural networks (DCNNs) always require a large number of labeled training data, which are not available for most medical image analysis applications due to the work required in image acquisition and particularly image annotation. In this paper, we propose a semi-supervised adversarial classification (SSAC) model that can be trained by using both labeled and unlabeled data for benign-malignant lung nodule classification. This model consists of an adversarial autoencoder-based unsupervised reconstruction network R, a supervised classification network C, and learnable transition layers that enable the adaption of the image representation ability learned by R to C. The SSAC model has been extended to the multi-view knowledge-based collaborative learning, aiming to employ three SSACs to characterize each nodule's overall appearance, heterogeneity in shape and texture, respectively, and to perform such characterization on nine planar views. The MK-SSAC model has been evaluated on the benchmark LIDC-IDRI dataset and achieves an accuracy of 92.53% and an AUC of 95.81%, which are superior to the performance of other lung nodule classification and semi-supervised learning approaches.

MicroRNA-422a functions as a tumor suppressor in non-small cell lung cancer through SULF2-mediated TGF-ß/SMAD signaling pathway.

MicroRNAs (miRNAs) have been demonstrated to participate in a variety of human cancers by functioning as post-transcriptional regulators of oncogenes or antioncogenes including non-small cell lung cancer (NSCLC). The aim of the current study was to identify the role of miR-422a in NSCLC via sulfatase 2 (SULF2) to further elucidate the mechanism of NSCLC. Initially, the expression of miR-422a and SULF2 was determined in NSCLC tissues and cells. The role of miR-422a in NSCLC was identified in relation with a miR-422a mimic or inhibitor, siRNA against SULF2 and TGF-ß1. The regulatory effects of miR-422a were examined following detection of the related epithelial mesenchymal transition (EMT)-related genes, and the apoptosis-related genes and evaluation of their cellular biological functions. The expression pattern of miR-422a, SULF2, and the TGF-ß/SMAD pathway-related genes was detected to elucidate the mechanism by which miR-422a influences the progression of NSCLC. Finally, xenograft tumors in nude mice were observed for tumorigenicity evaluation purposes. Our results showed that miR-422a was poorly expressed while SULF2 was highly expressed in NSCLC. Dual luciferase reporter gene assay further verified that miR-422a targeted SULF2. Altogether, this study demonstrated that miR-422a downregulated SULF2 to inhibit the TGF-ß/SMAD pathway. NSCLC cell proliferation, migration, invasion, colony formation, EMT and tumorigenesis were all inhibited while apoptosis was promoted upon restoration of miR-422a or silencing of SULF2. However, the activation of the TGF-ß/SMAD pathway was determined to reverse the tumor-suppressive effects of si-SULF2. miR-422a restoration, which ultimately inhibited the progression of NSCLC by suppressing the TGF-ß/SMAD pathway via SULF2.