Objective To observe the value of machine learning(ML)models based on brain functional network features combining clinical indicators for predicting postoperative outcomes of patients with drug-resistant mesial temporal lobe epilepsy(DR-mTLE).Methods Totally 84 patients with unilateral DR-mTLE who underwent surgery were retrospectively enrolled and classified into seizure free(SF)group(n=55)and non-seizure free(NSF)group(n=29)according to one-year postoperative follow-up.Clinical data were analyzed to screen independent predictors of postoperative outcomes.Based on brain preoperative resting-state functional MRI,brain functional networks were constructed using graph theory analysis,and 587 features were extracted.Five-fold cross validation was used to divide the data into training set and test set,then the optimal brain functional network features related to postoperative outcomes of DR-mTLE patients were selected.Combining with clinically relevant independent predictors,ML models were constructed using classifiers including Gaussian process(GP),logistic regression(LR),support vector machine(SVM)and quadratic discriminant analysis(QDA),respectively,and the prediction efficacy,calibration and clinical value of each ML model were evaluated.Results Both course of disease and lesion location were clinically relevant independent predictors of postoperative outcome of DR-mTLE patients(OR=0.928,5.710,P=0.010,0.016).Four optimal brain function network features were selected,including betweenness centrality of the third zone of cerebellar vermis,degree centrality of right globus pallidus,nodal efficiency of temporal left inferior temporal gyrus and nodal clustering coefficient of left inferior parietal lobule.The average area under the curve(AUC)of GP,LR,SVM and QDA models in test set was 0.868,0.864,0.875 and 0.870,respectively.Calibration curves and decision curve analysis indicated that each ML model had good calibration and high clinical net benefit.Conclusion ML models based on brain functional network features combining with clinical indicators could be used to effectively predict postoperative outcomes in DR-mTLE patients.

This research aimed to comprehensively understand the prevalence of mutation in drug-resistant molecular markers of imported Plasmodium vivax in Chongqing,the Pvmdr1,Pvdhps,Pvdhfr,Pvcrt-o and Pvk12 genes of Plasmodium vivax were systematically analyzed.Blood samples were collected from imported Plasmodium vivax-infected patients in Chongqing between 2011 and 2022.The Pvmdr1,Pvdhps,Pvdhfr,Pvcrt-o,and Pvk12 genes were amplified and then sequenced to precisely evaluate gene mutations.Bioinformatics methods were employed to conduct in depth analysis of the mutation prevalence.Regarding the Pvdhfr gene,mutations at codons 50,57,58,61,99,117,and 199 were detected in 2.9%,23.5%,76.4%,23.53%,2.9%,82.3%,and 5.88%of the samples,respectively.The double-mutant haplotype S58R/S117N was the most prevalent,accounting for 50%,followed by the quadruple-mutant haplotype F57L/S58R/T61M/S117T,which accounted for 11.76%.Among the four types of tandem-repeat variations of Pvdhfr,the wild-type was the most common,and the insertion type was a novel discovery in this study.For the Pvdhps gene,the prevalence among mutation genotypes was relatively low.The single-mutant genotype was dominant,constituting 27.03%.The prevalence of Pvmdr1 mutations at codons 958 and 1076 was 100%and 89.19%,respectively.Among the 37 successfully sequenced samples,K10 insertion was detected in only 8 cases(22.22%).Notably,no non-synonymous mutations of Pvk12 were identified in this study.The cases in this study were imported from various countries of origin.Novel tandem-repeat variation tyres of Pvdhfr and new mutation sites of Pvdhps were identifide,thus enriching the mutation information of imported Plasmodium vivax resistance molecular markers in China.

Objective To observe the value of machine learning(ML)models based on brain functional network features combining clinical indicators for predicting postoperative outcomes of patients with drug-resistant mesial temporal lobe epilepsy(DR-mTLE).Methods Totally 84 patients with unilateral DR-mTLE who underwent surgery were retrospectively enrolled and classified into seizure free(SF)group(n=55)and non-seizure free(NSF)group(n=29)according to one-year postoperative follow-up.Clinical data were analyzed to screen independent predictors of postoperative outcomes.Based on brain preoperative resting-state functional MRI,brain functional networks were constructed using graph theory analysis,and 587 features were extracted.Five-fold cross validation was used to divide the data into training set and test set,then the optimal brain functional network features related to postoperative outcomes of DR-mTLE patients were selected.Combining with clinically relevant independent predictors,ML models were constructed using classifiers including Gaussian process(GP),logistic regression(LR),support vector machine(SVM)and quadratic discriminant analysis(QDA),respectively,and the prediction efficacy,calibration and clinical value of each ML model were evaluated.Results Both course of disease and lesion location were clinically relevant independent predictors of postoperative outcome of DR-mTLE patients(OR=0.928,5.710,P=0.010,0.016).Four optimal brain function network features were selected,including betweenness centrality of the third zone of cerebellar vermis,degree centrality of right globus pallidus,nodal efficiency of temporal left inferior temporal gyrus and nodal clustering coefficient of left inferior parietal lobule.The average area under the curve(AUC)of GP,LR,SVM and QDA models in test set was 0.868,0.864,0.875 and 0.870,respectively.Calibration curves and decision curve analysis indicated that each ML model had good calibration and high clinical net benefit.Conclusion ML models based on brain functional network features combining with clinical indicators could be used to effectively predict postoperative outcomes in DR-mTLE patients.

This research aimed to comprehensively understand the prevalence of mutation in drug-resistant molecular markers of imported Plasmodium vivax in Chongqing,the Pvmdr1,Pvdhps,Pvdhfr,Pvcrt-o and Pvk12 genes of Plasmodium vivax were systematically analyzed.Blood samples were collected from imported Plasmodium vivax-infected patients in Chongqing between 2011 and 2022.The Pvmdr1,Pvdhps,Pvdhfr,Pvcrt-o,and Pvk12 genes were amplified and then sequenced to precisely evaluate gene mutations.Bioinformatics methods were employed to conduct in depth analysis of the mutation prevalence.Regarding the Pvdhfr gene,mutations at codons 50,57,58,61,99,117,and 199 were detected in 2.9%,23.5%,76.4%,23.53%,2.9%,82.3%,and 5.88%of the samples,respectively.The double-mutant haplotype S58R/S117N was the most prevalent,accounting for 50%,followed by the quadruple-mutant haplotype F57L/S58R/T61M/S117T,which accounted for 11.76%.Among the four types of tandem-repeat variations of Pvdhfr,the wild-type was the most common,and the insertion type was a novel discovery in this study.For the Pvdhps gene,the prevalence among mutation genotypes was relatively low.The single-mutant genotype was dominant,constituting 27.03%.The prevalence of Pvmdr1 mutations at codons 958 and 1076 was 100%and 89.19%,respectively.Among the 37 successfully sequenced samples,K10 insertion was detected in only 8 cases(22.22%).Notably,no non-synonymous mutations of Pvk12 were identified in this study.The cases in this study were imported from various countries of origin.Novel tandem-repeat variation tyres of Pvdhfr and new mutation sites of Pvdhps were identifide,thus enriching the mutation information of imported Plasmodium vivax resistance molecular markers in China.

Recent studies have characterized the genomic structures of many eukaryotic cells,often focusing on their relation to gene expression.However,these studies have largely investigated cells grown in 2D cultures,although the transcriptomes of 3D-cultured cells are generally closer to their in vivo phenotypes.To examine the effects of spatial constraints on chromosome conformation,we investigated the genomic architecture of mouse hepatocytes grown in 2D and 3D cultures using in situ Hi-C.Our results reveal significant differences in higher-order genomic interactions,notably in compartment identity and strength as well as in topologically associating domain(TAD)-TAD interactions,but only minor differences are found at the TAD level.Our RNA-seq analysis reveals up-regulated expression of genes involved in physiological hepatocyte functions in the 3D-cultured cells.These genes are associated with a subset of structural changes,suggesting that differences in genomic structure are critically important for transcriptional regulation.However,there are also many structural differences that are not directly associated with changes in gene expression,whose cause remains to be determined.Overall,our results indicate that growth in 3D significantly alters higher-order genomic interactions,which may be consequential for a subset of genes that are impor-tant for the physiological functioning of the cell.

Objective:To explore the clinical features and survival of newly diagnosed follicular lymphoma (FL) patients with diffuse large B-cell lymphoma (DLBCL) component.Methods:1845 newly diagnosed FL patients aged ≥ 18 years with grades 1-3a in 11 medical centers in China from 2000 to 2020 were included, and patients with DLBCL component were screened. The clinical data and survival data of the patients were retrospectively analyzed, and the prognostic factors were screened by univariate and multivariate analysis.Results:146 patients (7.9% ) with newly diagnosed FL had DLBCL component. The median age was 56 (25-83) years, 79 males (54.1% ) . The pathology of 127 patients showed the proportion of DLBCL component. Patients were divided into two groups according to whether the proportion of DLBCL component was ≥ 50% . The study found that patients with DLBCL component ≥ 50% had higher grade 3 ratio (94.3% vs 91.9% , P=0.010) , Ki-67 index ≥ 70% ratio (58.5% vs 32.9% , P=0.013) and PET-CT SUVmax ≥ 13 ratio (72.4% vs 46.3% , P=0.030) than patients with DLBCL component<50% . All patients received CHOP or CHOP like ± rituximab chemotherapy. The overall response rate (ORR) was 88.2% , and the complete response (CR) rate was 76.4% . In the groups with different proportions of DLBCL component, there was no significant difference in the remission rate after induction treatment and the incidence of disease progression within 2 years after initiation of treatment (POD24) ( P<0.05) . The overall estimated 5-year progression free survival (PFS) rate was 58.9% , and the 5-year overall survival (OS) rate was 90.4% . The 5-year OS rate of POD24 patients was lower than that of non POD24 patients (70.3% vs 98.5% , P<0.001) . Compared with non maintenance treatment of rituximab, maintenance treatment of rituximab could not benefit the 5-year PFS rate (57.7% vs 58.8% , P=0.543) , and the 5-year OS rate had a benefit trend, but the difference was not statistically significant (100% vs 87.8% , P=0.082) . Multivariate analysis showed that failure to reach CR after induction treatment was an independent risk factor for PFS ( P=0.006) , while LDH higher than normal was an independent risk factor for OS ( P=0.031) . Conclusion:FL patients with DLBCL component ≥50% have more invasive clinical and pathological features. CHOP/CHOP like ± rituximab regimen can improve the clinical efficacy of patients. Rituximab maintenance therapy can not benefit the PFS and OS of patients. Failure to reach CR after induction therapy was the independent unfavorable factor for PFS.

Objective:To investigate the occurrence of carotid artery abnormalities in human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) patients and the related risk factors affecting the occurrence of carotid artery abnormalities.Methods:A total of 169 regular follow-up outpatients with HIV/AIDS from April 2003 to December 2019 in Peking Union Medical College Hospital, whose carotid artery ultrasound examination were performed from July 2015 to December 2019 were included. The patients were divided into young (≤44 years old), middle-aged and elderly (≥45 years old) according to their ages, and the physical examination data of the included patients were collected.The statistical methods were logistic regression analysis and single sample t test. Results:Among the 169 HIV/AIDS patients, 40(23.7%) had abnormal carotid artery and 129(76.3%) had no abnormal carotid artery. Middle-aged and elderly people (odds ratio ( OR)=3.85, 95%confidence interval (95% CI) 1.54-9.65, P<0.01), hypertension ( OR=6.24, 95% CI 1.95-20.00, P<0.01), hyperlipidemia ( OR=2.44, 95% CI 1.00-5.93, P<0.05), and elevated human leucocyte antigen (HLA)-DR + CD8 + /CD8 + ( OR=1.03, 95% CI 1.01-1.06, P<0.05) were the risk factors for carotid artery abnormality. The common carotid artery inner medium film thickness (IMT) of patients in HIV/AIDS group Ⅰ (20 to 30 years old), group Ⅱ (31 to 40 years old), group Ⅲ (41 to 50 years old) were (0.061 0±0.001 2), (0.062 9±0.001 4) and (0.065 6±0.002 6) cm, respectively, which were thicker than the control groups ((0.051±0.003), (0.056±0.004) and (0.063±0.002) cm, respectively). The differences were all statistically significant ( t=5.119, 4.775 and 1.739, respectively, all P<0.05). The common carotid artery IMT of patients in HIV/AIDS group A (30 to 44 years old) and group B (45 to 59 years old) were (0.062 6±0.001 1) and (0.072 3±0.003 4) cm, respectively, which were thicker than the control groups ((0.052±0.011) and (0.064±0.015) cm, respectively), the differences were both statistically significant ( t=9.520 and 3.012, respectively, both P<0.01). Conclusion:Younger HIV-positive people have a higher probability of abnormal carotid arteries and may be at greater risk of cardiovascular disease than HIV-negative people of the same age, suggesting that HIV-positive people, especially young people, should be examined early with ultrasound of the neck arteries to detect abnormalities and intervene as soon as possible.

Objective: To investigate the characteristics and prognostic value of peripheral blood T lymphocyte subsets in patients with severe influenza. Methods: This was a single-center cross-sectional study in influenza patients admitted to Peking Union Medical College Hospital from August 2017 to April 2018. Peripheral blood lymphocyte subsets were detected by flow cytometry in both patients and 108 healthy controls. Influenza patients were divided into mild group and severe group. Severe patients were further classified into alive and fatal subgroups. Results: A total of 42 influenza patients were recruited in this study, including 24 severe cases (6 deaths). The remaining 18 cases were mild. The peripheral blood lymphocyte counts and lymphocyte subset counts (B, NK, CD4⁺T, CD8⁺T) in either mild patients[795 (571,1 007), 43 (23,144), 70 (47,135), 330 (256,457), 226 (148,366) cells/μl respectively] or severe patients[661 (474,1 151),92 (52,139), 54 (34,134), 373 (235,555), 180 (105,310) cells/μl respectively] were both significantly lower than those of healthy controls [1 963 (1 603,2 394),179 (119,239), 356 (231,496), 663 (531,824), 481 (341,693) cells/μl respectively]. Meanwhile, the T cells and CD8⁺T counts in fatal patients [370 (260,537) cells/μl and 87 (74,105) cells/μl] were significantly lower than those in severe and alive patients [722 (390,990) cells/μl and 222 (154,404) cells/μl]. CD8⁺HLA-DR/CD8⁺and CD8⁺CD38⁺/CD8⁺T cell activating subgroups in mild cases[(53.7±19.2)% and 74.8% (64.1%,83.7%) respectively] were significantly higher than those in severe cases[(38.5±21.7)% and 53.3% (45.3%,67.2%) respectively].Moreover,CD8⁺HLA-DR/CD8⁺count in severe and alive group was higher than that in fatal group [(46.1±19.1)% vs. (18.2±14.6)%, P<0.01]. Logistic regression analysis showed that CD8⁺T cell count (OR=0.952, 95%CI 0.910-0.997, P=0.035) and CD8⁺HLA-DR/CD8⁺T (OR=0.916, 95%CI 0.850-0.987, P=0.022) were both negatively correlated with mortality.Peripheral blood lymphocyte counts in mild cases rapidly decreased within 1 day after diagnosis, and returned to the basic level one week later. Conclusions All peripheral blood lymphocyte subsets (T,B,NK) in patients with influenza are significantly reduced. These findings are consistent with the immunological characteristics of respiratory viral infections, in which peripheral lymphocytes (especially T cells) migrate to respiratory tract in the early stage and circulate to the peripheral blood after recovery. The activated CD8⁺T cell counts in peripheral blood are negatively correlated with the severity of disease, which could be considered as a prognostic indicator of severe influenza.

Objective To investigate the clinical features and T lymphocytes subsets in patients with acquired immune deficiency syndrome (AIDS) and cytomegalovirus (CMV) infection.Methods A total of 48 hospitalized patients with human immunodeficiency virus (HIV)-1/AIDS and CMV infections were recruited at Peking Union Medical College Hospital from Jan 2010 to Aug 2017.Their clinical features and immune function were retrospectively analyzed.Patients with only HIV/AIDS in previous study were recruited as controls,Results All 48 patients were at C3 stage,including 36 men and 12 women.Five of them were younger than 30 years old,33 cases within 31-50 years old,and 10 cases older than 50 years old.Thirty-five patients had CD4+T lymphocytes ≤ 50 cells/μl,7 cases with CD4+T cells 51-100/μl,3 cases with 101-200 cells/μl,and 3 cases over 200 cells/μl.As to CMV infections,there were 31 cases of CMV viremia,1 case of CMV encephalitis,1 case of CMV enteritis,5 cases of CMV pneumonia,and 9 cases of CMV retinitis.Other opportunistic infections were also common including 16 cases of pneumocystis pneumonia,9 cases of tuberculosis,5 cases of syphilis,18 cases of digestive tract fungal infections,8 cases of pulmonary fungal infections,2 cases of EB virus infections,2 cases of HIV encephalopathy/progressive multifocal leukoencephalopathy (PML),3 cases of cryptococcal meningitis,1 case of toxoplasma infection.In group of both CMV and HIV/AIDS infections,100％ patients had inverted CD4+/CD8+ ratio.The immune activation marker CD8+CD38+/CD8+ was higher (61.6％-98.8％) with a median value of 91.2％ in 40 patients.HLA-DR+ CD8+/CD8+,another marker for T cell activation,was 25.5％-98.0％ in 44 patients with a median value of 60.3％.Thirty-six patients had both immune activation markers positive.There was no significant difference in counts of B cells,natural killer cells,CD4+ T cells,CD8+ T cells and immune activation subsets stratified by gender and age (P＞0.05).Meanwhile,neither serum HIV viral load nor serum CMV viral load was correlated with HLA-DR+CD8+/CD8+,CD8+CD38+/CD8+,CD4+T cell counts,and CD4+/CD8+ ratio in the CMV and HIV/AIDS co-infection group(all P＞0.05),while HIV viral load in HIV/AIDS only group was significantly correlated with HLA-DR+CD8+T/CD8+,CD38+CD8+/CD8+,CD4+ T cell counts,CD4+/CD8+ ratio (r=0.473,0.575,-0.767 and-0.678,respectively,all P＜0.05).Conclusions CMV infections develop in HIV patients with advanced stage.CMV infection can cause life-threatening multiple organ lesions,especially in those with CD4+ T cells less than 100 cells/μl.It is of great importance to screen CMV-IgM,pp65 antigen,CMV DNA to make early diagnosis and treatment.