A Point Mutation in Cassette Relieves the Repression Regulation of CcpA Resulting in an Increase in the Degradation of 2,3-Butanediol in Lactococcus lactis.

In lactic acid bacteria, the global transcriptional regulator CcpA regulates carbon metabolism by repressing and activating the central carbon metabolism pathway, thus decreasing or increasing the yield of certain metabolites to maximize carbon flow. However, there are no reports on the deregulation of the inhibitory effects of CcpA on the metabolism of secondary metabolites. In this study, we identified a single-base mutant strain of Lactococcus lactis N8-2 that is capable of metabolizing 2,3-butanediol. It has been established that CcpA dissociates from the catabolite responsive element (cre) site due to a mutation, leading to the activation of derepression and expression of the 2,3-butanediol dehydrogenase gene cluster (butB and butA). Transcriptome analysis and quantitative polymerase chain reaction (Q-PCR) results showed significant upregulation of transcription of butB and butA compared to the unmutated strain. Furthermore, micro-scale thermophoresis experiments confirmed that CcpA did not bind to the mutated cre. Furthermore, in a bacterial two-plasmid fluorescent hybridization system, it was similarly confirmed that the dissociation of CcpA from cre eliminated the repressive effect of CcpA on downstream genes. Finally, we investigated the differing catalytic capacities of the 2,3-butanediol dehydrogenase gene cluster in L. lactis N8-1 and L. lactis N8-2 for 2,3-butanediol. This led to increased expression of butB and butA, which were deregulated by CcpA repression. This is the first report on the elimination of the deterrent effect of CcpA in lactic acid bacteria, which changes the direction of enzymatic catalysis and alters the direction of carbon metabolism. This provides new perspectives and strategies for metabolizing 2,3-butanediol using bacteria in synthetic biology.

Engaging and (the Illusion of) Learning? Examining the Relationship Between Different Social Media Activities and Reproductive Health Knowledge.

Social media have become fundamental platforms for learning about health, including reproductive health knowledge. However, little is known about what specific user activity is conducive to learning about reproductive health and by what means. Drawing upon the cognitive mediation model, this study examines how different social media activities function in terms of elaboration and knowledge gain. Our hypothesized model was largely supported by a nationwide online survey with 1,000 Chinese women residing in both rural and urban areas. The results revealed the crucial role of information elaboration in bridging different social media activities with both subjective and factual reproductive health knowledge. Interestingly, public reposting of reproductive health information was found to be positively related to subjective knowledge but negatively related to factual knowledge, suggesting the emergence of an illusion of knowing among our participants. Multigroup SEM analyses revealed that the positive roles of scanning and private sharing in encouraging elaboration were more pronounced among users with lower levels of need for cognition. The findings are expected to contribute to a more nuanced understanding of health learning based on users' social media activities and intrinsic motivations for learning.

Alterations of functional brain activity and connectivity in female nurses working on long-term shift.

AIM: To investigate the alterations of functional brain activity and connectivity in female nurses working on long-term shifts and explore their correlations with work-related psychological traits. DESIGN: An exploratory cross-sectional study. METHODS: Thirty-five female nurses working on long-term shifts (shift nurses) and 35 female nurses working on fixed days (fixed nurses) were enrolled. After assessing the work-related psychological traits, including burnout, perceived stress, anxiety, and depression of nurses, the fractional amplitude of low-frequency fluctuations (fALFF) and region of interest (ROI)-based functional connectivity (FC) analyses were performed to investigate the differences of brain spontaneous activity and functional connectivity between these two groups of nurses. Thereafter, correlations between the functional brain parameters (fALFF and FC) and clinical metrics were investigated among the shift nurses. RESULTS: Compared to fixed nurses, shift nurses had higher burnout, perceived stress and depression scores, lower fALFF in the right dorsolateral prefrontal cortex (dlPFC), left and right superior parietal lobule (SPL), bilateral anterior cingulate cortex (ACC), and higher fALFF in the right superior/middle temporal gyrus, as well as decreased FC between the right dlPFC (the selected ROI) and bilateral ACC, left and right inferior frontal/orbitofrontal gyrus (IFG/IOFG), right SPL, and left middle occipital gyrus (voxel-level p < 0.001, cluster level p < 0.05, GRF correction). Correlation analyses demonstrated that the fALFF value of the right dlPFC was significantly correlated with the burnout and anxiety scores, the FC value of the right dlPFC-right SPL was correlated with the perceived stress and burnout scores, the FC value of the right dlPFC-right IFG/IOFG was correlated with the burnout score in shift nurses (p < 0.05). CONCLUSION: Shift nurses had work-related altered functional activity and connectivity in the right frontoparietal network, which provided objective and visualised evidence to clarify the hazards of long-term shift work on female nurses. PATIENT OR PUBLIC CONTRIBUTION: Seventy nurses participated deeply as subjects in this study. These findings are expected to draw managers' attention to the harmful influences of shift work on nurses.

Palladium-Catalyzed [3 + 2] Cycloaddition of 4-Vinyl-4-butyrolactones with Ketenes Generated in Situ from Acyl Chlorides: A Method for the Synthesis of Dihydrofurans.

In this paper, palladium-catalyzed [3 + 2] cycloaddition of 4-vinyl-4-butyrolactones with ketenes generated from easily available acyl chlorides was achieved. With Pd2(dba)3·CHCl3/XantPhos as the catalyst, the reaction proceeded smoothly under mild reaction conditions, affording a series of 2,3-dihydrofurans in moderate to high yields. The scale-up reaction and further transformations of the products are demonstrated, and a plausible mechanism is proposed as well.

T-type voltage-gated channels, Na+/Ca2+-exchanger, and calpain-2 promote photoreceptor cell death in inherited retinal degeneration.

Inherited retinal degenerations (IRDs) are a group of untreatable and commonly blinding diseases characterized by progressive photoreceptor loss. IRD pathology has been linked to an excessive activation of cyclic nucleotide-gated channels (CNGC) leading to Na+- and Ca2+-influx, subsequent activation of voltage-gated Ca2+-channels (VGCC), and further Ca2+ influx. However, a connection between excessive Ca2+ influx and photoreceptor loss has yet to be proven.Here, we used whole-retina and single-cell RNA-sequencing to compare gene expression between the rd1 mouse model for IRD and wild-type (wt) mice. Differentially expressed genes indicated links to several Ca2+-signalling related pathways. To explore these, rd1 and wt organotypic retinal explant cultures were treated with the intracellular Ca2+-chelator BAPTA-AM or inhibitors of different Ca2+-permeable channels, including CNGC, L-type VGCC, T-type VGCC, Ca2+-release-activated channel (CRAC), and Na+/Ca2+ exchanger (NCX). Moreover, we employed the novel compound NA-184 to selectively inhibit the Ca2+-dependent protease calpain-2. Effects on the retinal activity of poly(ADP-ribose) polymerase (PARP), sirtuin-type histone-deacetylase, calpains, as well as on activation of calpain-1, and - 2 were monitored, cell death was assessed via the TUNEL assay.While rd1 photoreceptor cell death was reduced by BAPTA-AM, Ca2+-channel blockers had divergent effects: While inhibition of T-type VGCC and NCX promoted survival, blocking CNGCs and CRACs did not. The treatment-related activity patterns of calpains and PARPs corresponded to the extent of cell death. Remarkably, sirtuin activity and calpain-1 activation were linked to photoreceptor protection, while calpain-2 activity was related to degeneration. In support of this finding, the calpain-2 inhibitor NA-184 protected rd1 photoreceptors.These results suggest that Ca2+ overload in rd1 photoreceptors may be triggered by T-type VGCCs and NCX. High Ca2+-levels likely suppress protective activity of calpain-1 and promote retinal degeneration via activation of calpain-2. Overall, our study details the complexity of Ca2+-signalling in photoreceptors and emphasizes the importance of targeting degenerative processes specifically to achieve a therapeutic benefit for IRDs. Video Abstract.

Large-Scale Molecular Dynamics Simulation Based on Heterogeneous Many-Core Architecture.

As the application of molecular dynamics (MD) simulations continues to evolve, the demand for accelerating large-scale simulation systems and handling of enormous simulation tasks is steadily increasing. We propose a parallel acceleration method for large-scale MD simulations based on Sunway heterogeneous many-core processors. This method integrates task scheduling, simulation calculations, and data storage, effectively tackling issues related to large-scale simulations and numerous simulation tasks. The task scheduling strategy flexibly handles tasks on various scales and enables parallel execution of multiple tasks. During the simulation calculations, we ported GROMACS to the Sunway architecture and accelerated the calculation of short-range forces through a heterogeneous processor. Our method achieves approximately 10-fold acceleration and 90% scalability when executing a single simulation task. When handling numerous simulation tasks, our method achieves parallel execution of all of the tasks with 90% scalability. By employing our method, we carried out 50 ns simulations on over 3000 distinct conotoxin structures individually within just 5 h. Additionally, we evaluated more than 200 protein-ligand complexes, and the simulation efficiency significantly exceeded that of midsized to small GPU clusters.

Sulphur-induced structural rearrangement in the self-sensitization photo-oxidation behaviour of phenothiazine-imidazole molecules.

A green photo-oxidation reaction was discovered in which the imidazole ring is oxidized and rearranged to generate imidazolinone derivatives with thermal activation delayed fluorescence properties. The excellent photo-oxidation properties of p-PTZ-PIM can be utilized for detecting trace oxygen in the water phase and the sealing of food packaging bags.

4-Vinylbenzodioxinones as a new type of precursor for palladium-catalyzed (4+3) cycloaddition of azomethine imines.

In this paper, benzo-fused cyclic carbonates were designed and synthesized as a new type of precursor of π-allylpalladium zwitterionic intermediates, and were applied in Pd-catalyzed diastereo- and enantioselective (4+3) cycloaddition with C,N-cyclic azomethine imines, leading to various biologically important 1,3,4-benzoxadiazepine derivatives in 43-99% yields with 6 : 1 to >20 : 1 dr and up to 95% ee.

Cu-Catalyzed Divergent Transformations of Allenylethylene Carbonates with Diboron Reagents.

Divergent transformations of allenylethylene carbonates with diboron reagents catalyzed by copper are disclosed. By using CuCl/IPr·HCl as the catalyst, the allenylethylene carbonates react with B2hex2 to afford 2,4-dien-1-ols as the product in the presence of Cs2CO3 as the base, iPrOH as the additive, and 1,4-dioxane as the solvent. And they react with B2pin2 to form boronic half acids in the presence of NaOtBu as the base, water as the additive, and THF as the solvent. The reactions afford corresponding products in good stereoselectivities and yields, and further derivatizations of boronic half acids and study of the mechanism are also demonstrated.

Performance of SHOX2 and RASSF1A methylation assay in supernatants and matched cell pellets for the diagnosis of malignant pleural effusion.

BACKGROUND: It is difficult to distinguish between malignant pleural effusion (MPE) and benign pleural effusion (BPE). The purpose of this study was to determine the best specimen type by evaluating the DNA methylation status of SHOX2 and RASSF1A in 3 matched PE components. METHODS: In total, 94 patients were enrolled, including 45 MPE, 35 BPE, and 14 undefined PE (UPE) with malignancies. PE samples were processed into supernatants, fresh-cell pellets, and formalin-fixed and paraffin-embedded (FFPE) cell blocks, respectively. A quantitative real-time PCR was used to detect the methylation status of SHOX2 and RASSF1A. RESULTS: SHOX2 and RASSF1A methylation levels were significantly higher in the 3 MPE sample types than those of BPE (P < 0.05). The area under the curve using cell-free DNA (cf-DNA) was the highest. The detection sensitivity of SHOX2 and RASSF1A in fresh-cell DNA, cf-DNA and FFPE cell-block were 71.1% (32/45), 97.8% (44/45) and 66.7% (28/42), respectively, with specificities of 97.1% (34/35), 94.3% (33/35), and 96.9% (31/32). Notably, a combination of the cytological analysis and cf-DNA methylation assay showed an increase in positivity rate from 75.6% to 100%. CONCLUSIONS: The SHOX2 and RASSF1A methylation assay using cf-DNA, the primary recommended specimen type, can excellently increase the diagnostic sensitivity of MPE. A combination of methylation assay with cytological analysis can be used for auxiliary diagnosis of PE.

Counteracting sexual and reproductive health misperceptions: Investigating the roles of stigma, misinformation exposure, and information overload.

OBJECTIVE: Sexual and reproductive health (SRH) misperceptions constitute a critical precursor to undesired health outcomes for women. Drawing on the model of stigma management communication and exposure effects, we aimed to investigate the underlying processes of SRH misperceptions. METHODS: A nationwide survey was conducted via quota sampling with Chinese women (N = 1000). Structural equation modeling with maximum likelihood estimation and 5000 bootstrapping resamples were used to test the hypotheses. RESULTS: Stigma perceptions positively predicted information avoidance (ß = 0.207, p < 0.001), which, in turn, was positively associated with misperceptions (ß = 0.195, p < 0.001). Misinformation exposure significantly predicted misperceptions (ß = 0.607, p < 0.001), and this relationship was mediated by information avoidance (ß = 0.020, 95% CI [0.007, 0.040]). Moreover, information overload strengthened the relationship between misinformation exposure and information avoidance (ß = 0.153, p < 0.001) as well as the relationship between misinformation exposure and misperceptions (ß = 0.077, p = 0.006). CONCLUSION: Stigma and misinformation exposure play prominent roles in the formation of SRH misperceptions. Information overload facilitates the misinformation-misperception transformation. PRACTICE IMPLICATIONS: To counteract SRH misperceptions, health education should alleviate SRH stigma perceptions and strategically design messages to avoid information avoidance and overload.

Molecular generation strategy and optimization based on A2C reinforcement learning in de novo drug design.

MOTIVATION: In the field of pharmacochemistry, it is a time-consuming and expensive process for the new drug development. The existing drug design methods face a significant challenge in terms of generation efficiency and quality. RESULTS: In this paper, we proposed a novel molecular generation strategy and optimization based on A2C reinforcement learning. In molecular generation strategy, we adopted transformer-DNN to retain the scaffolds advantages, while accounting for the generated molecules' similarity and internal diversity by dynamic parameter adjustment, further improving the overall quality of molecule generation. In molecular optimization, we introduced heterogeneous parallel supercomputing for large-scale molecular docking based on message passing interface communication technology to rapidly obtain bioactive information, thereby enhancing the efficiency of drug design. Experiments show that our model can generate high-quality molecules with multi-objective properties at a high generation efficiency, with effectiveness and novelty close to 100%. Moreover, we used our method to assist shandong university school of pharmacy to find several candidate drugs molecules of anti-PEDV. AVAILABILITY AND IMPLEMENTATION: The datasets involved in this method and the source code are freely available to academic users at https://github.com/wq-sunshine/MomdTDSRL.git.

Pregnancy outcomes following natural conception and assisted reproduction treatment in women who received COVID-19 vaccination prior to conception: a population-based cohort study in China.

Introduction: The coronavirus disease-2019 (COVID-19) pandemic has swept across the world and continues to exert serious adverse effects on vulnerable populations, including pregnant women and neonates. The vaccines available at present were designed to prevent infection from COVID-19 strains and control viral spread. Although the incidence of pregnancy cycle outcomes are not likely to increase patients vaccinated prior to pregnancy compared with unvaccinated patients based on our knowledge of vaccination safety, there is no specific evidence to support this hypothesis. Therefore, the current study aimed to investigate the association between maternal vaccination prior to conception and pregnancy outcomes. Methods: We retrospectively analyzed 2,614 women who received prenatal care and delivered in the Obstetrical Department of The First Affiliated Hospital of Anhui Medical University between February 2022 and November 2022. Of the 1,380 eligible pregnant women, 899 women who had received preconception vaccination were assigned to a vaccine group and 481 women who were not vaccinated were control group. Of the enrolled patients, 291 women received fertility treatment (141 vaccinated women, 150 unvaccinated women). The primary outcomes were pregnancy complications (hypothyroidism, gestational diabetes mellitus, pregnancy-induced hypertension, polyhydramnios, oligohydramnios, premature rupture of membranes and postpartum hemorrhage), obstetric outcomes (preterm birth rate, cesarean section rate) and neonatal outcomes (birth-weight, body length, low-birth-weight rate, rate of congenital defects, neonatal mortality and admission to the neonatal intensive care unit). Results: There was no significant difference in the incidence of complications during pregnancy and delivery when compared between the vaccine group and control group in either univariate- or multivariate-models. The type of vaccine was not associated with the odds of adverse pregnancy outcome. Among the women with infertility treatment, the vaccinated group and the unvaccinated group had similar pregnancy outcomes. Conclusion: Women who received COVID-19 vaccination prior to conception had similar maternal and neonatal outcomes as women who were unvaccinated. Our findings indicate that COVID-19 vaccinations can be safely administered prior to pregnancy in women who are planning pregnancy or assisted reproductive treatment. During new waves of COVID-19 infection, women who are planning pregnancy should be vaccinated as soon as possible to avoid subsequent infections.

ZNF143 inhibits hepatocyte mitophagy and promotes non-alcoholic fatty liver disease by targeting increased lncRNA NEAT1 expression to activate ROCK2 pathway.

BACKGROUND: Nonalcoholic fatty liver disease (NAFLD) is the most common hepatic disorders worldwide. The mitophagy is suggested to be repressed in NAFLD, but the mechanism remains to be elucidated. METHODS: NAFLD cell and mouse models were established by treating with free fatty acid (FFA) and feeding a high fat diet (HFD), respectively. QRT-PCR, Western blotting, or IHC measured the expression of ZNF143, lncRNA NEAT1, ROCK2, and lipid formation/mitophagy-related proteins. Cell viability and mitophagy were evaluated by MTT and immunofluorescence. The chloroform-methanol extraction method measured triglyceride and total cholesterol levels. ELISA detected ALT and AST levels. The interactions among ZNF143, lncRNA NEAT1 and SND1 were analysed by ChIP, dual-luciferase reporter, pull-down, and RIP. The lipid droplets were determined by Oil-red O and HE staining. RESULTS: ZNF143 and lncRNA NEAT1 were upregulated in hepatic cells treated with FFA (p < 0.01 and p < 0.001). Knockdown of ZNF143 or lncRNA NEAT1 inhibited lipid droplets formation, while promoting mitophagy (p < 0.01 and p < 0.001). ZNF143 promoted lncRNA NEAT1 transcriptional expression through binding to its promoter. LncRNA NEAT1 increased ROCK2 mRNA stability by targeting SND1. LncRNA NEAT1 or ROCK2 overexpression reversed the effect of ZNF143 or lncRNA NEAT1 knockdown on hepatic steatosis and mitophagy (p < 0.01 and p < 0.001). ZNF143 or lncRNA NEAT1 knockdown inhibited HFD-induced steatosis and promoted mitophagy in vivo (p < 0.01 and p < 0.001). CONCLUSION: The upregulation of lncRNA NEAT1 caused by ZNF143 promoted NAFLD through inhibiting mitophagy via activating ROCK2 pathway by targeting SND1, providing potential targets for NAFLD therapy.

Palladium-Catalyzed Asymmetric (3 + 2) Cycloaddition of 5-Allenyloxazolidine-2,4-Diones with Barbiturate-Derived Alkenes: Synthesis of Spirobarbiturate-γ-Lactams.

The 5-allenyloxazolidine-2,4-diones had been synthesized as novel precursors of π-allyl palladium zwitterion and were applied in a palladium-catalyzed enantioselective (3 + 2) annulation by using barbiturate-derived alkenes as the reaction partner in the presence of an axially chiral phosphoramidite ligand. This reaction proceeded smoothly under mild reaction conditions, affording highly functionalized spirobarbiturate-γ-lactam derivatives in excellent yields along with high diastereo- and enantioselectivities. The scale-up reaction and further transformation of the product were also successful.

Soluble polymer facilely self-grown in situ on conducting substrates at room temperature towards electrochromic applications.

Electrochromic polymer film preparation methods such as spin coating, spray coating, and electrochemical polymerization, are commonly used. At present, developing new film preparation technology is an important aspect in the field of electrochromics. Herein, a continuous in situ self-growing method based on the chemical reaction occurring on the surface of an ITO glass between a metal oxide and organic acid groups was successfully applied to prepare electrochromic polymer films at a mild room temperature. SEM, FT-IR spectroscopy, XPS, and XRD characterization methods were combined to reveal the process and mechanism of film formation. The following notable electrochromic properties were observed: switching time within 6 s, contrast reached 35%, and minimal decrease of stability after 600 cycles. Finally, the patterned films were obtained through the directional growth of polymers in solution. This study provides an effective strategy for designing and preparing electrochromic films by self-growing methods in future applications.

Pathological complete response to long-course neoadjuvant alectinib in lung adenocarcinoma with EML4-ALK rearrangement: report of two cases and systematic review of case reports.

Objective: Despite the promising efficacy and tolerability of alectinib in treating advanced anaplastic lymphoma kinase (ALK) positive non-small cell lung cancer (NSCLC), the role of alectinib in neoadjuvant setting remains understudied in ALK-rearranged resectable lung cancer. Methods: Our report concerns two cases of early-stage NSCLC with complete pathologic responses to off-label use of long-course neoadjuvant alectinib. PubMed, Web of Science, and Cochrane Library were searched comprehensively for ALK-positive resectable cases with neoadjuvant alectinib. The papers were chosen following PRISMA recommendations. Seven cases from the literature and two present cases were evaluated. Results: Two cases with stage IIB (cT3N0M0) EML4-ALK lung adenocarcinoma received long-course (more than 30 weeks) of neoadjuvant alectinib followed by R0 lobectomy with the complete pathological response. In our systematic review, 74 studies were included in the original search. Application of the screening criteria resulted in 18 articles deemed eligible for full-text reading. Following the application of the exclusion criteria, out of six papers, seven cases were selected for inclusion in the final analysis and were included in the systematic review. None of the studies were included in the quantitative analysis. Conclusion: We report two cases of lung adenocarcinoma with resectable ALK-positive that achieved pCR with long-course neoadjuvant alectinib. Our cases and a systematic review of the literature support the feasibility of neoadjuvant alectinib treatment for NSCLC. However, large clinical trials must be conducted in the future to determine the treatment course and efficacy of the neoadjuvant alectinib modality. Systematic review registration: https://www.crd.york.ac.uk/PROSPERO, identifier CRD42022376804.

HDAC inhibition delays photoreceptor loss in Pde6b mutant mice of retinitis pigmentosa: insights from scRNA-seq and CUT&Tag.

Purpose: This research aimed to ascertain the neuroprotective effect of histone deacetylase (HDAC) inhibition on retinal photoreceptors in Pde6brd1 mice, a model of retinitis pigmentosa (RP). Methods: Single-cell RNA-sequencing (scRNA-seq) explored HDAC and poly (ADP-ribose) polymerase (PARP)-related gene expression in both Pde6b-mutant rd1 and wild-type (WT) mice. The CUT&Tag method was employed to examine the functions of HDAC in rd1 mice. Organotypic retinal explant cultures from WT and rd1 mice were exposed to the HDAC inhibitor SAHA (suberoylanilide hydroxamic acid) postnatally, from day 5 to day 11. The terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) assay was applied to quantify the percentage of photoreceptor loss in the outer nuclear layer (ONL). HDAC activity was confirmed to be inhibited by SAHA through an HDAC activity assay. Moreover, the study evaluated PARP activity, a key driver of the initial response to DNA damage during photoreceptor degeneration, following HDAC inhibition. Results: The scRNA-seq revealed that diverse roles of HDAC and PARP isoforms in photoreceptor cell death. HDAC-related genes appeared to regulate cell death and primary immunodeficiency. Alterations in HDAC activity were consistent with the TUNEL-positive cells in the ONL at different time points. Notably, SAHA significantly postponed photoreceptor loss and decreased HDAC and PARP activity, thereby implicating both in the same degenerative pathway. Conclusions: This study highlights that the interaction between HDAC inhibition and PARP can delay photoreceptor cell death, proposing a promising therapeutic approach for RP.

Deep learning-based diagnosis of histopathological patterns for invasive non-mucinous lung adenocarcinoma using semantic segmentation.

OBJECTIVES: The application of artificial intelligence (AI) to the field of pathology has facilitated the development of digital pathology, hence, making AI-assisted diagnosis possible. Due to the variety of lung cancers and the subjectivity of manual evaluation, invasive non-mucinous lung adenocarcinoma (ADC) is difficult to diagnose. We aim to offer a deep learning solution that automatically classifies invasive non-mucinous lung ADC histological subtypes. DESIGN: For this investigation, 523 whole-slide images (WSIs) were obtained. We divided 376 of the WSIs at random for model training. According to WHO diagnostic criteria, six histological components of invasive non-mucinous lung ADC, comprising lepidic, papillary, acinar, solid, micropapillary and cribriform arrangements, were annotated at the pixel level and employed as the predicting target. We constructed the deep learning model using DeepLab v3, and used 27 WSIs for model validation and the remaining 120 WSIs for testing. The predictions were analysed by senior pathologists. RESULTS: The model could accurately predict the predominant subtype and the majority of minor subtypes and has achieved good performance. Except for acinar, the area under the curve of the model was larger than 0.8 for all the subtypes. Meanwhile, the model was able to generate pathological reports. The NDCG scores were greater than 75%. Through the analysis of feature maps and incidents of model misdiagnosis, we discovered that the deep learning model was consistent with the thought process of pathologists and revealed better performance in recognising minor lesions. CONCLUSIONS: The findings of the deep learning model for predicting the major and minor subtypes of invasive non-mucinous lung ADC are favourable. Its appearance and sensitivity to tiny lesions can be of great assistance to pathologists.