AUC Values Research Articles

Comparative Study of Potential Habitats for Two Endemic Grassland Caterpillars on the Qinghai-Tibet Plateau Based on BIOMOD2 and Land Use Data.

This study has systematically investigated and compared the geographical distribution patterns and population density of G. menyuanensis (Gm) and G. qinghaiensis (Gq), which are endemic to the QTP region and inflict severe damage. Using a method combining the BIOMOD2 integration model (incorporating nine ecological niche models) and current species distribution data, this study has compared changes in potential habitats and distribution centers of these two species during ancient, present, and future climate periods and conducted a correlation test on the prediction results with land use types. The study results indicate that there are differences in geographical distribution patterns, distribution elevations, and population density of these two species. Compared with single models, the integration model exhibits prominent accuracy and stability with higher KAPPA, TSS, and AUC values. The distribution of suitable habitats for these two species is significantly affected by climatic temperature and precipitation. There is a significant difference between the potential habitats of these two species. Gm and Gq are distributed in the northeastern boundary area and the central and eastern areas of the QTP, respectively. The areas of their suitable habitats are significantly and positively correlated with the area of grassland among all land use types of QTP, with no correlations with the areas of other land use types of QTP. The potential habitats of both species during the paleoclimate period were located in the eastern and southeastern boundary areas of the QTP. During the paleoclimate period, their potential habitats expanded towards the Hengduan Mountains (low-latitude regions) in the south compared with their current suitable habitats. With the subsequent temperature rising, their distribution centers shifted towards the northeast (high-latitude) regions, which could validate the hypothesis that the Hengduan Mountains were refuges for these species during the glacial period. In the future, there will be more potential suitable habitats for these two species in the QTP. This study elucidates the ecological factors affecting the current distribution of these grass caterpillars, provides an important reference for designating the prevention and control areas for Gm and Gq, and helps protect the alpine meadow ecosystem in the region.

Machine learning-based prognostic prediction for hospitalized HIV/AIDS patients with cryptococcus infection in Guangxi, China

ObjectiveTo develop and validate a machine learning model for predicting mortality-associated prognostic factors in order to reduce in-hospital mortality rates among HIV/AIDS patients with Cryptococcus infection in Guangxi, China.MethodsThis retrospective prognostic study included HIV/AIDS patients with cryptococcosis in the Fourth People’s Hospital of Nanning from October 2011 to June 2019. Clinical features were extracted and used to train ten machine learning models, including Logistic Regression, KNN, DT, RF, Adaboost, Xgboost, LightGBM, Catboost, SVM, and NBM, to predict the outcome of HIV patients with cryptococcosis infection. The sensitivity, specificity, AUC, and F1 value were applied to assess model performance in both the testing and training sets. The optimal model was selected and interpreted.ResultsA total of 396 patients were included in the study. The average in-hospital mortality of HIV/AIDS patients with cryptococcosis was 12.9% from 2012 to 2019. After feature screening, 20 clinical features were selected for model construction, accounting for 93.8%, including ART, Electrolyte disorder, Anemia, and 17 laboratory tests. The RF model (AUC 0.9787, Sensitivity 0.9535, Specificity 0.8889, F1 0.7455) and the SVM model (AUC 0.9286, Sensitivity 0.7907, Specificity 0.9786, F1 0.8293) had excellent performance. The SHAP analysis showed that the primary risk factors for prognosis prediction were identified as BUN/CREA, Electrolyte disorder, NEUT%, Urea, and IBIL.ConclusionsRF and SVM machine learning models have shown promising predictive abilities for the prognosis of hospitalized HIV/AIDS patients with cryptococcosis, which can aid clinical assessment and treatment decisions for patient prognosis.

Comparative analysis of scoring systems for predicting mortality in Fournier gangrene: single center, 15years experience.

To compare 5 published scoring systems (Fournier gangrene severity index [FGSI], Uludag FGSI [UFGSI], age-adjusted Charlson Comorbidity Index [ACCI] and the Quick Sequential Organ Failure Assessment (qSOFA) and the Laboratory Risk Indicator for Necrotizing fasciitis (LRINEC) scores and to evaluate risk factors for outcome prediction in patients with Fournier gangrene (FG). Between 2010 and 2024, 311 patients were included in the study. The data of 276 survivors and 55 non-survivors were compared. Univariate and multivariate analyses were performed to determine which of the 5 scoring systems calculated according to demographic, laboratory and clinical data predicted mortality better. In addition, a cut-off value for these scoring systems was determined by ROC analysis. The mortality rate was 17.6% in FG patients. In all 5 scoring systems, univariate analyses showed higher values in non-survivor patients, while UFGSI and FGSI scoring systems were superior to other scoring systems in multivariate analyses (p < 0.001). ROC analysis using mortality-based sensitivity and specificity revealed that the optimal cut-off values for FGSI, UFGSI, ACCI, SOFA and LRINEC should be equal to or higher than 10, 12, 4, 2 and 8, respectively. UFGSI and FGSI scores had the highest AUC values among all scores. This was followed by ACCI, qSOFAand LRINEC. For a UFGSI score ≥ 12, the sensitivity was 84%, specificity 97%, PPV 85% and NPV 97%. Among the scoring systems used to predict mortality, UFGSI was the most accurate, while LRINEC was the least accurate.

Does the Combined Arteritis Damage Score (CARDS) in the Diagnosis of Takayasu Arteritis Predict Prognosis and Need for Biologic Therapy?

We aimed to investigate the prognostic value of the combined arteritis damage score (CARDS) in Takayasu arteritis (TAK) patients to predict the need for biologic treatment at diagnosis and the possible contribution of wall thickness (WT). Blind evaluation of MRA/CTA at the time of diagnosis was performed by a reader rheumatologist (RR) and an interventional radiologist (RIR). The CARDS damage score for 21 arterial regions was assessed as normal, mild or moderate/severe stenosis, occclusion or aneursym/dilatation. Additionally, WT was scored for all regions as present or absent. A modified CARDS (mCARDS) was calculated as the sum of CARDS and the number of WT areas. According to follow-up treatment, 10 patients with non-biologic treatment (non-BT) (F/M:8/2, median age 37.5years) and 15 patients with biologic treatment (BT) (F/M:13/2, median age 30years) were included. Indian Takatasu Arteritis Score (ITAS), CRP, and ESR levels were similar in both groups. CARDS (1.4 (0-7.2) vs 4.5 (.6-19), P: .003), WT (1.5 (0-8) vs 7 (1-21), P < .001), and mCARDS (4 (0-14.2) vs 11.4 (1.6-40), P < .001) scores were significantly higher in the BT group compared to nonBT group. Cohen's kappa coefficient between RR and RIR for WT was .99 with 99.6% aggrement, and CARDS was .98 with 99.6% agreement. The AUC values for CARDS, WT, and mCARDS scores were .748 (.605-.892), .837 (.723-.950), and .847 (.735-.958), respectively, and P value was <.0001. The prediction of prognosis and biologic treatment need at TAK diagnosis using non-invasive angiographic images can improve outcomes and prompt closer follow-up. The combination of CARDS and WT as mCARDS achieved the highest sensitivity and specificity, and all scores appear useful for predicting prognosis.

A machine learning approach to predict foot care self-management in older adults with diabetes

BackgroundFoot care self-management is underutilized in older adults and diabetic foot ulcers are more common in older adults. It is important to identify predictors of foot care self-management in older adults with diabetes in order to identify and support vulnerable groups. This study aimed to identify predictors of foot care self-management in older adults with diabetes using a machine learning approach.MethodThis cross-sectional study was conducted between November 2023 and February 2024. The data were collected in the endocrinology and metabolic diseases departments of three hospitals in Turkey. Patient identification form and the Foot Care Scale for Older Diabetics (FCS-OD) were used for data collection. Gradient boosting algorithms were used to predict the variable importance. Three machine learning algorithms were used in the study: XGBoost, LightGBM and Random Forest. The algorithms were used to predict patients with a score below or above the mean FCS-OD score.ResultsXGBoost had the best performance (AUC: 0.7469). The common predictors of the models were age (0.0534), gender (0.0038), perceived health status (0.0218), and treatment regimen (0.0027). The XGBoost model, which had the highest AUC value, also identified income level (0.0055) and A1c (0.0020) as predictors of the FCS-OD score.ConclusionThe study identified age, gender, perceived health status, treatment regimen, income level and A1c as predictors of foot care self-management in older adults with diabetes. Attention should be given to improving foot care self-management among this vulnerable group.

Prediction and validation of pathologic complete response for locally advanced rectal cancer under neoadjuvant chemoradiotherapy based on a novel predictor using interpretable machine learning

BackgroundPrecise evaluation of pathological complete response (pCR) is essential for determining the prognosis of patients with locally advanced rectal cancer (LARC) undergoing neoadjuvant chemoradiotherapy (NCRT) and can offer clues for the selection of subsequent treatment strategies. Most current predictive models for pCR focus primarily on pre-treatment factors, neglecting the dynamic systemic changes that occur during neoadjuvant chemoradiotherapy, and are constrained by low accuracy and lack of integrity. PurposeThis study devised a novel predictor of pCR using dynamic alterations in systemic inflammation-nutritional marker indexes (SINI) during neoadjuvant therapy and developed a machine-learning model to predict pCR. MethodsTwo cohorts of patients with LARC from center one from 2012 to 2017 and from center two from 2020 to 2023 were integrated for analysis. This study compared dynamic changes in blood indexes before and after neoadjuvant therapy and surgical operation. A least absolute shrinkage and selection operator (LASSO) regression analysis was conducted to mitigate collinearity and identify key indexes, constructing the SINI. Univariate and multiple logistic regression analyses were used to identify the independent risk factors associated with pCR. Additionally, 10 machine learning algorithms were employed to develop predictive models to assess risk. The hyperparameters of the machine learning models were optimized using a random search and 10-fold cross-validation. The models were assessed by examining various metrics, including the area under the receiver operating characteristic curves (AUC), the area under the precision-recall curve (AUPRC), decision curve analysis, calibration curves, and the precision and accuracy of the internal and external validation cohorts. Additionally, Shapley's additive explanations (SHAP) were employed to interpret the machine learning models. ResultsThe study cohort comprised 677 patients from the center one and 224 patients from the center two. Six key indexes were identified, and a predictive index, SINI, was constructed. Univariate and multiple logistic regression analyses revealed that SINI, clinical T-stage, clinical N-stage, tumor size, and the distance from the anal verge were independent risk factors for pCR in patients with LARC following NCRT. The mean AUC value of the extreme gradient boosting (XGB) model in the 10-fold cross-validation of the training set was 0.877. The XGB model demonstrated superior performance in the internal and external validation sets. Specifically, in the internal test set, the XGB model achieved an AUC of 0.86, AUPRC of 0.707, accuracy of 0.82, and precision of 0.80. In the external validation set, the XGB model exhibited an AUC of 0.83, AUPRC of 0.702, accuracy of 0.81, and precision of 0.81. Additionally, the predictions generated by the XGB model were analyzed using SHAP. ConclusionThis study involved developing and validating an XGB model using SINI to predict pCR in patients with LARC. Besides, a SINI-based machine learning model shows promise in accurately predicting pCR following NCRT in patients with resectable LARC, offering valuable insights for personalized treatment approaches.

7515 Plasma proteomic signatures for type 2 diabetes mellitus and related traits in the UK Biobank

Abstract Disclosure: T. Gupte: None. Z. Azizi: None. K. Nzenkue: None. P. Kho: None. M. Chen: None. J. Zhou: None. R. Guarischi-Sousa: None. D.J. Panyard: None. F. Abbasi: None. K. Watson: None. S. Clarke: None. P.S. Tsao: None. T.L. Assimes: None. Background: The plasma proteome holds promise as a future diagnostic and prognostic tool that can accurately reflect complex human traits and disease processes. We assessed the ability of plasma proteins to associate with two primary risk factors of type 2 diabetes mellitus (T2DM) and to predict incident T2DM. Methods: Clinical, genetic, and high-throughput proteomic data from three subcohorts of 5586, 5584, and 36339 individuals in the UK Biobank were analyzed for association with DXA-measured truncal fat, maximum VO2, and incident T2DM, respectively. Exclusion criteria included prevalent T2DM, history of T1DM, or elevated A1c (&amp;gt;48 mmol/mol) values. We used LASSO to assess relative ability of 13 clinical variables including the Q-Diabetes (Q-D) score, 36 polygenic risk scores (PRSs), and 2923 proteins to associate with each trait by comparing variances explained (R2) and AUC statistics between data types. Stability selection with randomized LASSO identified the most robustly associated proteins for each trait. The benefit of proteomic signatures (PSs) over the Q-D score was evaluated through the derivation of Δ AUC values. Separate case-control groups were developed to evaluate T2DM prediction in the adiposity and fitness cohorts. Results: Across all three subcohorts, the mean age was 56.6 years, 54.9% were female, and 5.4% developed T2DM over a median follow-up of 7.7 years. In the adiposity cohort, a protein-only LASSO model explained 73.2% of the variance in truncal fat and GLIPR1, CLMP, LEP, FABP4, and FAM3C were among the top proteins selected. In the fitness cohort, a protein-only LASSO model explained 61.9% of the variance in maximum VO2 and LEP, CD99, AGRN, ELN, and CA14 were among the top proteins selected. LASSO-derived PSs increased the R2 of truncal fat and maximum VO2 over clinical and genetic factors by 13% and 1%, respectively, with the former also improving risk prediction of T2DM over Q-D [Δ AUC: 0.026 (95% CI -0.007, 0.060)]. In the T2DM cohort, a protein-only LASSO model returned an AUC of 0.856 (95% CI 0.840, 0.871) and PLXNB2, PODXL, PLA2G7, FGFR2, and NFASC were among the top proteins selected. We observed a similar improvement in T2DM prediction over Q-D [Δ AUC: 0.015 (95% CI 0.009, 0.023)] when using a PS derived strictly from the T2DM outcome and augmented with non-overlapping proteins associated with fat and fitness. A few proteins (ranging between 15-33 for each trait) identified by stability selection algorithms offered most of the improvement in risk prediction. Conclusion: Plasma PS modestly improves the prediction of T2DM over that possible with clinical and genetic factors. Further studies are warranted to better elucidate the clinical utility of these signatures in predicting T2DM risk over the standard practice of using the Q-D score. Presentation: 6/2/2024

Comparison of Omega-3 polyunsaturated fatty acids bioavailability in fish oil and krill oil: Network Meta-analyses

Background and Aims: Fish oil and krill oil are primary sources of Omega-3 polyunsaturated fatty acids (PUFAs), differing in their molecular forms and bioavailability. Understanding these differences can optimize their therapeutic use. This study aims to compare the bioavailability of Omega-3 PUFAs in fish oil and krill oil using a network meta-analysis approach. By evaluating various molecular forms and dosages, the study seeks to identify the most effective Omega-3 sources and dosing regimens to maximize health benefits.Methods: The study adhered to the explanation of the PRISMA network meta-analysis 2015 (registered PROSPERO ID: CRD42024532536). The risk of bias was evaluated using the PEDro scale. This network meta-analysis incorporated data from studies published between 2003 and 2023, sourced from five databases: ClinicalTrials.gov, PubMed, Cochrane CENTRAL, the International Clinical Trials Registry Platform-World Health Organization, and Embase. Statistical analysis was conducted using advanced models in R software to ensure a rigorous evaluation of bioavailability.Results: Out of 26 high-quality studies, findings reveal superior bioavailability of krill oil compared to fish oil. Specifically, fish oil above 3000 mg, in re-esterified triacylglycerol or ethyl ester formulations (100–2900 mg), and krill oil (100–1900 mg) significantly enhanced the Omega-3 index. At lower dosages (under 2000 mg), krill oil shows superior Omega-3 absorption compared to fish oil. Using fish oil ethyl ester at doses between 2000 and 2900 mg may be the most effective for reducing Tmax. However, this finding should be interpreted with caution due to high heterogeneity and limited statistical significance. The highest AUC values were observed in krill oil phospholipid/free fatty acid formulation. Emulsion forms of fish oil are more effective in increasing Cmax than other formulations. These results were supported by robust statistical evidence (Egger's test, p > 0.05), highlighting the effectiveness of specific formulations and doses in optimizing Omega-3 absorption.Conclusion: The research highlights the importance of understanding the bioavailability of Omega-3 PUFAs from fish oil and krill oil. The findings suggest that low-dose krill oil and fish oil emulsions (under 2000 mg) are effective and potentially safer alternatives to high-dose fish oil, potentially minimizing the risk of adverse effects.

Exploring the prognostic and diagnostic value of lactylation-related genes in sepsis

The discovery of Lactylation may pave the way for novel approaches to investigating sepsis. This study focused on the prognostic and diagnostic significance of lactylated genes in sepsis. RNA sequencing was performed on blood samples from 20 sepsis patients and 10 healthy individuals at Southwest Medical University in Luzhou, Sichuan, China. Genes associated with sepsis were identified through analysis of RNA sequencing data. Afterward, the genes that were expressed differently were compared with the lactylation genes, resulting in the identification of 55 lactylation genes linked to sepsis. The overlapping genes underwent analysis using Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). Protein-Protein Network Interactions were used to screen for the core genes. The datasets GSE65682, GSE69528, GSE54514, GSE63042, and GSE95233 were obtained from the GEO database to validate core genes. Survival analysis evaluated the predictive significance of central genes in sepsis, while Receiver Operating Characteristic (ROC) Curve analysis was employed to establish the diagnostic value of genes. Additionally, a meta-analysis was conducted to confirm the precision of RNA-seq data. We obtained five peripheral blood samples, including two from healthy individuals, one from a patient with systemic inflammatory response syndrome (SIRS), and two from patients with sepsis. These samples were used to identify the specific location of core genes using 10×single-cell sequencing. High-throughput sequencing and bioinformatics techniques identified two lactylation-related genes (S100A11 and CCNA2) associated with sepsis. Survival analysis indicated that septic patients with reduced levels of S100A11 had a decreased 28-day survival rate compared to those with elevated levels. Conversely, individuals exhibiting decreased CCNA2 levels demonstrated a greater likelihood of surviving for 28 days than those in the high expression category, indicating a favorable association with survival rates among sepsis patients (P < 0.05). Both genes showed high sensitivity and specificity based on the ROC curve, with AUC values of 0.961 for S100A11 and 0.890 for CCNA2. The meta-analysis revealed that S100A11 exhibited high levels of expression in the sepsis survivors, whereas it displayed low levels of expression in the non-survivors; on the other hand, CCNA2 demonstrated low expression in the sepsis survivors and high expression in the non-survivors (P < 0.05). Single-cell RNA sequencing ultimately showed that monocyte macrophages, T cells, and B cells exhibited high expression levels of the crucial genes associated with sepsis-induced lactylation. In conclusion, the lactylation genes S100A11 and CCNA2 are strongly linked to sepsis and could be valuable markers for diagnosing, predicting outcomes, and providing guidance for sepsis.

Proteomics Analysis of Renal Cell Line Caki-2 with AFMID Overexpression and Potential Biomarker Discovery in Urine.

Aromatic caninurine formamase (AFMID) is an enzyme involved in the tryptophan pathway, metabolizing N-formylkynurenine to kynurenine. AFMID had been found significantly downregulated in clear cell renal cell carcinoma (ccRCC) in both tissue and urine samples. Although ccRCC is characterized by a typical Warburg-like phenotype, mitochondrial dysfunction, and elevated fat deposition, it is unknown whether AFMID plays a role in tumorigenesis and the development of ccRCC. In the present study, AFMID overexpression had inhibitory effects for ccRCC cells, decreasing the rate of cell proliferation. Quantitative proteomics showed that AFMID overexpression altered cellular signaling pathways involved in cell growth and cellular metabolism pathways, including lipid metabolism and inositol phosphate metabolism. Further urine proteomic analysis indicated that cellular function dysfunction with AFMID overexpression could be reflected in the urine. The activity of predicted upregulators DDX58, TREX1, TGFB1, SMARCA4, and TNF in ccRCC cells and urine showed opposing change trends. Potential urinary biomarkers were tentatively discovered and further validated using an independent cohort. The protein panel of APOC3, UMOD, and CILP achieved an AUC value of 0.862 for the training cohort and 0.883 for the validation cohort. The present study is of significance in terms of highlighting various aspects of pathway changes associated with AFMID enzymes, discovering potential specific biomarkers for potential patient diagnosis, and therapeutic targeting.

A generalizable and interpretable model for early warning of pest-induced crop diseases using environmental data

Pest infestations and resulting crop diseases threaten global food security. Traditional pest and disease monitoring methods are time-consuming and prone to delays, thus necessitating the development of effective prediction strategies that facilitate early and timely detection. In response to this challenge, this research proposes a generalizable and interpretable machine learning model to predict two major rice pests—Green Leafhopper and Yellow Stem Borer—and a No Pest class using environmental data collected from various regions in India. The dataset, comprising factors such as temperature, humidity, and rainfall, underwent rigorous preprocessing. Key features like temperature difference, humidity difference, and vapor pressure deficit were engineered to enhance the model’s performance. Twelve baseline models were trained and their performance was evaluated using F1 scores and AUC values due to the imbalanced nature of the dataset. Through statistical analysis of baseline models, Random Forest, Balanced Random Forest, XGBoost, and CatBoost models, were selected for hyperparameter tuning via Optuna. The tuned CatBoost model demonstrated superior performance, achieving AUC values of 0.99 for Green Leafhoppers, 0.98 for Yellow Stem Borers, 1.00 for No Pest class and an overall F1 score of 0.9414 with a mean AUC value of 0.9912 across all classes. Additionally, Explainable Artificial Intelligence techniques, particularly SHAP, were employed to interpret the model’s predictions, identifying the relative importance of environmental factors in pest occurrence. This interpretability aligns the model’s predictions with established agronomic knowledge, enhancing its practical utility for early pest and disease detection. The generalizability of the proposed model suggests it can be adapted to other crops and regions, offering a valuable tool for early warning systems in agriculture, promoting sustainable practices, and reducing crop losses due to pests and diseases.

NAS FD Lung: A novel lung assist diagnostic system based on neural architecture search

In the detection and recognition of lung nodules, pulmonary nodules vary in size and shape and contain many similar tissues and organs around them, leading to the problems of both missed detection and false detection in existing detection algorithms. Designing proprietary detection and recognition networks manually requires substantial professional expertise. This process is time-consuming and labour-intensive and leads to issues like parameter redundancy and improper feature selection. Therefore, this paper proposes a new pulmonary CAD (computer-aided diagnosis) system for pulmonary nodules, NAS FD Lung (Using the NAS approach to search deep FPN and DPN networks), that can automatically learn and generate a deep learning network tailored to pulmonary nodule detection and recognition task requirements. NAS FD Lung aims to use automatic search to generate deep learning networks in the auxiliary diagnosis of pulmonary nodules to replace the manual design of deep learning networks. NAS FD Lung comprises two automatic search networks: BM NAS-FPN (Using NAS methods to search for deep FPN structures with Binary operation and Matrix multiplication fusion methods) network for nodule detection and NAS-A-DPN (Using the NAS approach to search deep DPN networks with attention mechanism) for nodule identification. The proposed technique is tested on the LUNA16 dataset, and the experimental results show that the model is superior to many existing state-of-the-art approaches. The detection accuracy of lung nodules is 98.23%. Regarding the lung nodules classification, the accuracy, specificity, sensitivity, and AUC values achieved 96.32%,97.14%,95.82%, and 98.33%, respectively.

Evaluation of lymphovascular space invasion in endometrial carcinoma by APTw and mDixon-Quant.

Lymphovascular space invasion (LVSI) is a strong and independent risk factor that increases the probability of endometrial carcinoma (EC) recurrence and reduces the survival rate of patients. To investigate the value of amide proton transfer weighted (APTw) and mDixon-Quant techniques in evaluating EC lymphovascular space invasion (LVSI). Data of 50 EC patients (18 LVSI+ and 32 LVSI-) confirmed by surgery and pathology were retrospectively analyzed. Preoperative magnetic resonance imaging (MRI) scans included APTw and mDixon-Quant imaging. APT, transverse relaxation rate (R2*), and fat fraction (FF) plots were obtained by postprocessing. The APT, R2*, and FF values of the two groups of cases were measured by two observers. The agreement between the two observers was good. The mean APT, R2*, and FF values of LVSI+ EC were 2.947% ± 0.399%, 20.605 /s (range = 18.525-27.953), and 2.234% ± 1.047%, respectively, while the parameters of LVSI- EC were 2.628% ± 0.307%, 18.968 /s (range = 16.225-20.544), and 2.103% ± 1.070%, respectively. The APT and R2* values of LVSI+ EC were higher than those of LVSI- EC (P < 0.05). There was no significant difference in FF value between the two groups. The AUC values of APT, R2*, and APT + R2* for LVSI were 0.751, 0.713, and 0.781, respectively (all P > 0.05). APT value was moderately correlated with R2* value (r = 0.528, P < 0.001) and weakly correlated with FF value (r = 0.312, P = 0.027). APTw and mDixon-Quant techniques could evaluate the LVSI status of EC, and their combined application could improve diagnostic efficiency.

B-124 Structure-based Variant Effect Prediction and Design of Functional CYP2C9 Variants by Virtual Evolution

Abstract Background The decryption of massive variants of uncertain significance (VUS) discovered by sequencing poses the main challenge in the post-sequencing era. VUS also resulted in a particular problem in the field of pharmacogenomics, which can be epitomized by drug-metabolizing enzymes, such as CYP2C9. Therefore, an accurate model that incorporates the inner connection between massive protein variants and their function alterations is crucial for clinical annotation of human variome and the virtual evolution of enzyme engineering for biotherapeutics. Methods We developed a model named variant effect recognition network for CYP2C9 (vERnet-P), which learned features from AlphaFold2-predicted protein structures and predicted the enzyme activity of missense single-nucleotide variants in CYP2C9. The two crucial strategies of vERnet-P are the construction of amino acid interaction networks and the application of deep learning. Therefore, vERnet-P leveraged highly accurate protein structures predicted by AlphaFold2, combined with novel techniques for enriching and capturing useful information. Based on the accurate and preemptive prediction of CYP2C9 variants, we performed the saturation mutation prediction on several sites of CYP2C9 to identify new variants with the expected function. Furthermore, the drug metabolic activities of these newly identified variants were investigated with two probe drugs by in vitro metabolic activity assessments. Results An accuracy of up to 93.5% for the classification of activity levels was yielded in the testing dataset. In addition, the tasks for recognizing the high-activity and low-activity variants both achieved 93.5% accuracies, demonstrating that vERnet-P truly learned the features related to enzyme activity, rather than the selection bias to one kind of samples. Strikingly, the AUC values of the ROC and the PR reached 0.971 and 0.966 respectively in the testing cohort for vERnet-P. vERnet-P achieved the strongest agreement with massively parallel assessments relative to the other 8 computational variant effect predictors in the task of predicting CYP2C9 variant enzyme activity. We explored 6 mutant sites in CYP2C9 by using vERnet-P to perform saturation mutation prediction and found 12 novel variants with a high possibility of activity changing. We identified 6 variants with most likely increased activity by using the prediction of wild-type as a cutoff, consequently, 6 variants with most likely decreased activity were selected. According to these 12 discovered variants, 10 were strongly confirmed by the in vitro metabolic activity assays of both probe drugs. Conclusions vERnet-P achieved state-of-the-art performance in predicting the CYP2C9 variant enzyme activity. The preemptive prediction for CYP2C9 variants can efficiently provide the clinical drug dosing guideline. Furthermore, by saturating mutation prediction on several sites, our results indicated the evolutionary direction of CYP2C9 enzyme activity and discovered some brand-new CYP2C9 variants with ultra-strong functional alterations, which have not previously been reported. The results of in vitro metabolic activity assessment were highly consistent with our in silico prediction, indicating the potential of AI models for inferring evolutionary direction and designing novel variants with the expected function.

Immunogenic cell death signatures from on-treatment tumor specimens predict immune checkpoint therapy response in metastatic melanoma

Melanoma is a highly malignant form of skin cancer that typically originates from abnormal melanocytes. Despite significant advances in treating metastatic melanoma with immune checkpoint blockade (ICB) therapy, a substantial number of patients do not respond to this treatment and face risks of recurrence and metastasis. This study collected data from multiple datasets, including cohorts from Riaz et al., Gide et al., MGH, and Abril-Rodriguez et al., focusing on on-treatment samples during ICB therapy. We used the single-sample gene set enrichment analysis (ssGSEA) method to calculate immunogenic cell death scores (ICDS) and employed an elastic network algorithm to construct a model predicting ICB efficacy. By analyzing 18 ICD gene signatures, we identified 9 key ICD gene signatures that effectively predict ICB treatment response for on-treatment metastatic melanoma specimens. Results showed that patients with high ICD scores had significantly higher response rates to ICB therapy compared to those with low ICD scores. ROC analysis demonstrated that the AUC values for both the training and validation sets were around 0.8, indicating good predictive performance. Additionally, survival analysis revealed that patients with high ICD scores had longer progression-free survival (PFS). This study used an elastic network algorithm to identify 9 ICD gene signatures related to the immune response in metastatic melanoma. These gene features can not only predict the efficacy of ICB therapy but also provide references for clinical decision-making. The results indicate that ICD plays an important role in metastatic melanoma immunotherapy and that expressing ICD signatures can more accurately predict ICB treatment response and prognosis for on-treatment metastatic melanoma specimens, thus providing a basis for personalized treatment.

Developing a rapid predictive model for falls in older hospitalized patients.

This study was aimed to identify the independent risk factors for falls n hospitalized older patients and develop a corresponding predictive model. A retrospective observational study design was adopted, comprising 440 older patients with falls history and 510 older patients without falls history during hospitalization. Data collected included demographic information, vital signs, comorbidities, psychiatric disorder, function absent, current medication, other clinical indicators. Mobility disability, high-risk medications use, frequency of hospitalizations, psychiatric disorder, visual impairment are independent risk factors for falls in older patients. The A-M2-HPV scoring system was developed. The AUC value of the nomogram was 0.884, indicating the model has excellent discriminative ability. The AUC value of the A-M2-HPV score was 0.788, demonstrating better discrimination and stratification capabilities. The A-M2-HPV scoring system provides a valuable tool to assess the risk of falls in hospitalized older patients and to aid in the implementation of preventive measures.

A-249 Monocyte distribution width (MDW) as a reliable diagnostic biomarker for sepsis in HIV-infected patients

Abstract Background Sepsis is the main cause of death in patients with acquired immunodeficiency syndrome (AIDS) and early diagnosis of sepsis in AIDS remains a significant challenge. Monocyte distribution width (MDW), as a new type of sepsis auxiliary diagnostic marker, has attracted widespread attention in the research and exploration of its clinical application. This study aims to evaluate the diagnostic efficacy of MDW in the occurrence of sepsis in HIV-infected patients and to construct a new clinical diagnostic model in combination MDW with other indicators. Methods A prospective cohort study was conducted at Shanghai Public Health Center with a total of 602 patients enrolled between December 2022 and August 2023. Sepsis-3 criteria were employed as the diagnostic standard for sepsis. MDW was measured at admission, using the hematology analyzer DxH900 (Beckman Coulter Inc.). Comprehensive clinical data, including complete blood counts and biochemical profiles, were collected. Enrolled participants received a follow-up about their survivals on day 30 and day 90. Five machine learning models were evaluated and compared to select the best performing model. Results The diagnostic performance of MDW for sepsis yielded an AUC of 0.818. With a cut-off value of 25.26, the sensitivity and specificity reached 0.822 and 0.754 respectively. The best performing variable among all five ML algorithms was Random Forest model. The top three variables with the highest importance ranking were platelet, albumin, and hemoglobin. Incorporating MDW and the top three laboratory indicators into a random forest model, the AUC value improved to 0.931. With a cut-off value of 0.5, the sensitivity was 1.00 while the specificity was 0.732. Conclusions MDW was an auxiliary diagnostic biomarker for sepsis in HIV-infected patients with reliable sensitivity and specificity. In combination MDW with other common laboratory biomarkers, a diagnostic model can be established to further improve the diagnostic efficiency of sepsis.

Improving predictions of rock tunnel squeezing with ensemble Q-learning and online Markov chain

Predicting rock tunnel squeezing in underground projects is challenging due to its intricate and unpredictable nature. This study proposes an innovative approach to enhance the accuracy and reliability of tunnel squeezing prediction. The proposed method combines ensemble learning techniques with Q-learning and online Markov chain integration. A deep learning model is trained on a comprehensive database comprising tunnel parameters including diameter (D), burial depth (H), support stiffness (K), and tunneling quality index (Q). Multiple deep learning models are trained concurrently, leveraging ensemble learning to capture diverse patterns and improve prediction performance. Integration of the Q-learning-Online Markov Chain further refines predictions. The online Markov chain analyzes historical sequences of tunnel parameters and squeezing class transitions, establishing transition probabilities between different squeezing classes. The Q-learning algorithm optimizes decision-making by learning the optimal policy for transitioning between tunnel states. The proposed model is evaluated using a dataset from various tunnel construction projects, assessing performance through metrics like accuracy, precision, recall, and F1-score. Results demonstrate the efficiency of the ensemble deep learning model combined with Q-learning-Online Markov Chain in predicting surrounding rock tunnel squeezing. This approach offers insights into parameter interrelationships and dynamic squeezing characteristics, enabling proactive planning and support measures implementation to mitigate tunnel squeezing hazards and ensure underground structure safety. Experimental results show the model achieves a prediction accuracy of 98.11%, surpassing individual CNN and RNN models, with an AUC value of 0.98.

Prediction of PD-L1 expression in unresectable hepatocellular carcinoma with gadoxetic acid-enhanced MRI

ObjectivesTo develop a model to predict programmed death-ligand 1 (PD-L1) expression in unresectable hepatocellular carcinoma (HCC) based on gadoxetic acid-enhanced magnetic resonance imaging (MRI) findings and clinical characteristics. Materials and methodsWe enrolled patients with unresectable HCC who underwent gadoxetic acid-enhanced MRI between January 2021 and May 2023. Immunohistochemical staining of PD-L1 was performed on a biopsy specimen. Patients with a history of any prior treatment for HCC or those lacking an MRI scan within 30 days of the biopsy date were excluded. Using the clinical and MRI findings, we developed a PD-L1 prediction score using logistic regression. ResultsThis study included 49 patients with HCC (median age, 64 years; interquartile range, 57–73 years; 44 men). Among these, 15 (31 %) were positive for PD-L1 expression. The PD-L1 prediction score was defined as the sum of arterial phase hypoenhancement (score 1), necrosis (score 1), and AFP >4000 ng/mL (score 2). The AUC value of the PD-L1 prediction score was 0.838 (95 % confidence interval [CI], 0.715–0.962). When the PD-L1 prediction score was ≥3, the sensitivity, specificity, and positive predictive value of PD-L1 positivity were 67 %, 91 %, and 77 %, respectively. ConclusionWe developed a PD-L1 prediction score for unresectable HCC with high specificity that could potentially contribute to the identification of effective candidates for immune checkpoint inhibitors.

Identification of 10 differentially expressed genes involved in the tumorigenesis of cervical cancer via next-generation sequencing.

The incidence and mortality of cervical cancer remain high in female malignant tumors worldwide. There is still a lack of diagnostic and prognostic markers for cervical carcinoma. This study aimed to screen differentially expressed genes (DEGs) between normal and cervical cancer tissues to identify candidate genes for further research. Uterine cervical specimens were resected from our clinical patients after radical hysterectomy. Three patients' transcriptomic datasets were built by the next generation sequencing (NGS) results. DEGs were selected through the edgeR and DESeq2 packages in the R environment. Functional enrichment analysis, including GO/DisGeNET/KEGG/Reactome enrichment analysis, was performed. Normal and cervical cancer tissue data from the public databases TCGA and GTEx were collected to compare the expression levels of 10 selected DEGs in tumor and normal tissues. ROC curve and survival analysis were performed to compare the diagnostic and prognostic values of each gene. The expression levels of candidate genes were verified in 15 paired clinical specimens via quantitative real-time polymerase chain reaction. There were 875 up-regulated and 1,482 down-regulated genes in cervical cancer samples compared with the paired adjacent normal cervical tissues according to the NGS analysis. The top 10 DEGs included APOD, MASP1, ACKR1, C1QTNF7, SFRP4, HSPB6, GSTM5, IGFBP6, F10 and DCN. GO, DisGeNET and Reactome analyses revealed that the DEGs were related to extracellular matrix and angiogenesis which might influence tumorigenesis. KEGG enrichment showed that PI3K-Akt signaling pathway might be involved in cervical cancer tumorigenesis and progression. The expression levels of selected genes were decreased in tumors in both the public database and our experimental clinical specimens. All the candidate genes showed excellent diagnostic value, and the AUC values exceeded 0.90. Additionally, APOD, ACKR1 and SFRP4 expression levels could help predict the prognosis of patients with cervical cancer. In this study, we selected the top 10 DEGs which were down-regulated in cervical cancer tissues. All of them had dramatically diagnostic value. APOD, ACKR1 and SFRP4 were associated with the survivals of cervical cancer. C1QTNF7, HSPB6, GSTM5, IGFBP6 and F10 were first reported to be candidate genes of cervical carcinoma.