Supervised learning of enhancer-promoter specificity based on genome-wide perturbation studies highlights areas for improvement in learning.

MOTIVATION: Understanding the rules that govern enhancer-driven transcription remains a central unsolved problem in genomics. Now with multiple massively parallel enhancer perturbation assays published, there are enough data that we can utilize to learn to predict enhancer-promoter (EP) relationships in a data-driven manner. RESULTS: We applied machine learning to one of the largest enhancer perturbation studies integrated with transcription factor (TF) and histone modification ChIP-seq. The results uncovered a discrepancy in the prediction of genome-wide data compared to data from targeted experiments. Relative strength of contact was important for prediction, confirming the basic principle of EP regulation. Novel features such as the density of the enhancers/promoters in the genomic region was found to be important, highlighting our lack of understanding on how other elements in the region contribute to the regulation. Several TF peaks were identified that improved the prediction by identifying the negatives and reducing False Positives. In summary, integrating genomic assays with enhancer perturbation studies increased the accuracy of the model, and provided novel insights into the understanding of enhancer-driven transcription. AVAILABILITY AND IMPLEMENTATION: The trained models, data, and the source code are available at http://doi.org/10.5281/zenodo.11290386 and https://github.com/HanLabUNLV/sleps.

Comparison of Speech Outcomes Between Speech Therapy Only and Double-Opposing Z-Plasty Combined With Speech Therapy in Patients With Submucous Cleft Palate.

The purpose of this study was to compare speech outcomes in patients with submucous cleft palate (SMCP) between speech therapy alone and double-opposing Z-plasty (DOZ) combined with speech therapy. The subjects were 67 patients with SMCP (overt type, 45 males, 22 females), who were divided into the observation group (n=18), the speech therapy group (n=24; duration, 17.8 mo), and the DOZ and speech therapy (DOZ-speech therapy) group (n=25; median age at DOZ, 5.3 years, duration, 18.6 mo). The median age at initial and final speech assessments were 3 and 5 years. After age, sex, syndromic status, duration of speech therapy, surgery timing, and speech outcomes were investigated, statistical analysis was performed. After tailored interventions, both isolated and non-isolated SMCP patients experienced significant improvements in speech outcomes, including nasal emission, hypernasality, compensatory articulation, and unintelligible speech. Since comparable improvements were observed, there were no significant differences in the final assessments regardless of initial speech issues between the speech therapy group and the DOZ-speech therapy group (all P>0.05). In the DOZ-speech therapy group, the rate of achieving "socially acceptable" speech was 92.3% in isolated cases and 90% in non-isolated cases. Multivariate analysis revealed that DOZ showed a tendency to reduce hypernasality, compensatory articulation, and "unintelligible" speech; syndromic or developmental conditions influenced outcomes in nasal emission and hypernasality; and initial hypernasality and compensatory articulation were correlated with outcomes. Therefore, DOZ surgery could be recommended to resolve hypernasality and compensatory articulation in SMCP patients before speech issues worsen.

Intraoperative neuromonitoring during thyroidectomy does not decrease vocal cord palsy risk, but the cumulative experience of the surgeon may.

PURPOSE: To evaluate the influence of intraoperative neuromonitoring (IONM) on vocal cord palsy (VCP) rates and assess the temporal trends in VCP rates. METHODS: The subjects of this retrospective study were patients who underwent thyroidectomy for thyroid cancer between March, 2014 and June, 2022, at a university hospital in Korea. We compared VCP rates between the non-IONM and IONM groups and analyzed the risk factors for VCP and VCP rates over time. RESULTS: A total of 712 patients were included in the analysis. The rates of transient and permanent VCP did not differ significantly between the non-IONM and IONM groups. Transient VCP occurred in 4.6% and 4.3% patients (p = 0.878) and VCP was permanent in 0.7% and 0.4% patients (p = 0.607) in the non-IONM and IONM groups, respectively. Among the nerves at risk, transient damage occurred in 2.8% and 3.0% patients (p = 0.901), and permanent damage occurred in 0.4% and 0.3% (p = 0.688), respectively. Multivariate analysis revealed no significant risk factors for VCP. There was a significant decreasing trend in VCP rates over time as the cumulative number of cases increased (p = 0.017). CONCLUSIONS: IONM did not reduce the risk of VCP significantly. However, the declining trend of VCP rates suggests that the surgeon's experience may mitigate VCP risk.

Quality of Life and Nutritional Outcomes of Stomach-Preserving Surgery for Early Gastric Cancer: A Secondary Analysis of the SENORITA Randomized Clinical Trial.

Importance: The Sentinel Node Oriented Tailored Approach (SENORITA) randomized clinical trial evaluated quality of life (QoL) and nutritional outcomes between the laparoscopic sentinel node navigation surgery (LSNNS) and laparoscopic standard gastrectomy (LSG). However, there has been no report on the QoL and nutritional outcomes of patients who underwent stomach-preserving surgery among the LSNNS group. Objective: To compare long-term QoL and nutritional outcomes between patients who underwent stomach-preserving surgery and those who underwent standard gastrectomy and to identify factors associated with poor QoL outcomes in patients who underwent stomach-preserving surgery. Design, Setting, and Participants: This study is a secondary analysis of the SENORITA trial, a randomized clinical trial comparing LSNNS with LSG. Patients from 7 tertiary or general hospitals across the Republic of Korea were enrolled from March 2013 to December 2016, with follow-up through 5 years. Data were analyzed between August and September 2022. Among trial participants, patients who underwent actual laparoscopic standard gastrectomy in the LSG group and those who underwent stomach-preserving surgery in the LSNNS group were included. Patients who did not complete the baseline or any follow-up questionnaire were excluded. Intervention: Stomach-preserving surgery vs standard gastrectomy. Main Outcomes and Measures: Overall European Organization for Research and Treatment of Cancer QoL Questionnaire Core 30 (EORTC QLQ-C30) and stomach module (STO22) scores, body mass index, hemoglobin, protein, and albumin levels. Results: A total of 194 and 257 patients who underwent stomach-preserving surgery and standard gastrectomy, respectively, were included in this study (mean [SD] age, 55.6 [10.6] years; 249 [55.2%] male). The stomach-preserving group had better QoL scores at 3 months postoperatively in terms of physical function (87.2 vs 83.9), dyspnea (5.9 vs 11.2), appetite loss (13.1 vs 19.4), dysphagia (8.0 vs 12.7), eating restriction (10.9 vs 18.2), anxiety (29.0 vs 35.2), taste change (7.4 vs 13.0), and body image (19.5 vs 27.2). At 1 year postoperatively, the stomach-preserving group had significantly higher body mass index (23.9 vs 22.1, calculated as weight in kilograms divided by height in meters squared) and hemoglobin (14.3 vs 13.3 g/dL), albumin (4.3 vs 4.25 g/dL), and protein (7.3 vs 7.1 g/dL) levels compared to the standard group. Multivariable analyses showed that tumor location (greater curvature, lower third) was favorably associated with global health status (ß, 10.5; 95% CI, 3.2 to 17.8), reflux (ß, -8.4; 95% CI, -14.7 to -2.1), and eating restriction (ß, -5.7; 95% CI, -10.3 to -1.0) at 3 months postoperatively in the stomach-preserving group. Segmental resection was associated with risk of diarrhea (ß, 40.6; 95% CI, 3.1 to 78.1) and eating restriction (ß, 15.1; 95% CI, 1.1 to 29.1) at 3 years postoperatively. Conclusions and Relevance: Stomach-preserving surgery after sentinel node evaluation was associated with better long-term QoL and nutritional outcomes than standard gastrectomy. These findings may help facilitate decision-making regarding treatment for patients with early-stage gastric cancer. Trial Registration: ClinicalTrials.gov Identifier: NCT01804998.

A comparison of RNA-Seq data preprocessing pipelines for transcriptomic predictions across independent studies.

BACKGROUND: RNA sequencing combined with machine learning techniques has provided a modern approach to the molecular classification of cancer. Class predictors, reflecting the disease class, can be constructed for known tissue types using the gene expression measurements extracted from cancer patients. One challenge of current cancer predictors is that they often have suboptimal performance estimates when integrating molecular datasets generated from different labs. Often, the quality of the data is variable, procured differently, and contains unwanted noise hampering the ability of a predictive model to extract useful information. Data preprocessing methods can be applied in attempts to reduce these systematic variations and harmonize the datasets before they are used to build a machine learning model for resolving tissue of origins. RESULTS: We aimed to investigate the impact of data preprocessing steps-focusing on normalization, batch effect correction, and data scaling-through trial and comparison. Our goal was to improve the cross-study predictions of tissue of origin for common cancers on large-scale RNA-Seq datasets derived from thousands of patients and over a dozen tumor types. The results showed that the choice of data preprocessing operations affected the performance of the associated classifier models constructed for tissue of origin predictions in cancer. CONCLUSION: By using TCGA as a training set and applying data preprocessing methods, we demonstrated that batch effect correction improved performance measured by weighted F1-score in resolving tissue of origin against an independent GTEx test dataset. On the other hand, the use of data preprocessing operations worsened classification performance when the independent test dataset was aggregated from separate studies in ICGC and GEO. Therefore, based on our findings with these publicly available large-scale RNA-Seq datasets, the application of data preprocessing techniques to a machine learning pipeline is not always appropriate.

Real-world Nationwide Outcomes of Minimally Invasive Surgery for Advanced Gastric Cancer Based on Korean Gastric Cancer Association-Led Survey.

PURPOSE: The study aimed to investigate real-world surgical outcomes of minimally invasive surgery (MIS) for advanced gastric cancer using Korean Gastric Cancer Association (KGCA)-led nationwide data. MATERIALS AND METHODS: A nationwide survey of patients who underwent surgical treatment for gastric cancer in 2019 was conducted by the KGCA. A total of 14,076 patients from 68 institutions underwent surgery, and 4,953 patients diagnosed with pathological stages IB-III gastric cancer were included. Among them, 1,689 patients who underwent MIS (MIS group) and 1,689 who underwent the open approach (open group) were matched using propensity score in a 1:1 ratio. Surgical outcomes were compared, and multivariate analysis was performed to identify the independent factors for overall morbidity. RESULTS: The MIS group had a lower proportion of D2 lymphadenectomy, total omentectomy, and combined resection. However, the number of harvested lymph nodes was higher in the MIS group. Better surgical outcomes, including less blood loss and shorter hospital stay, were observed in the MIS group, and the overall morbidity rate was significantly lower in the MIS group (17.5% vs. 21.9%, P=0.001). The mortality rates did not differ significantly between the 2 groups. In the multivariate analysis, the minimally invasive approach was a significant protective factor against overall morbidity (odds ratio, 0.799; P=0.006). CONCLUSIONS: Based on the Korean nationwide data, MIS for stage IB-III gastric cancer had better short-term outcomes than the open approach, including lower rates of wound complications, intra-abdominal abscesses, and cardiac problems.

Clinical Efficacy of Laparoscopic Sentinel Node Navigation Surgery for Stomach Preservation in Patients With Early Gastric Cancer: 5-year Results of the SENORITA Trial.

OBJECTIVE: This study aimed to compare laparoscopic standard gastrectomy (LSG) and laparoscopic sentinel node navigation surgery (LSNNS) for EGC in terms of 5-year long-term oncologic outcomes. SUMMARY BACKGROUND DATA: The oncological safety of LSNNS for early gastric cancer (EGC) has not been confirmed. Three-year disease-free survival (DFS), which is the primary endpoint of the phase III multicenter randomized controlled clinical trial (SEntinel Node ORIented Tailored Approach [SENORITA] trial), did not show the non-inferiority of LSNNS relative to LSG. METHODS: The SENORITA trial, a multicenter randomized clinical trial, was designed to show that LSNNS is non-inferior to LSG in terms of 3-year DFS. In the present study, we collected 5-year follow-up data from 527 patients recruited in the SENORITA trial as the full analysis set (FAS). Disease-free survival (DFS), overall survival (OS), disease-specific survival (DSS), and recurrence patterns were evaluated using the FAS of both LSG (n=269) and LSNNS (n=258). RESULTS: The 5-year DFS was not significantly different between the LSG and LSNNS groups (P=0.0561). During the 5-year follow-up, gastric cancer-related events, such as metachronous cancer, were more frequent in the LSNNS group than in the LSG group. However, ten recurrent cancers in the remnant stomach of both groups were curatively resected by additional gastrectomy and one by additional endoscopic resection. Two of the 198 patients who underwent local resection for stomach preservation based on the LSNNS results developed distant metastasis. However, there was no statistically significant difference in the 5-year OS and DSS (P=0.7403 and P=0.9586, respectively) between the two groups. CONCLUSION: The 5-year DFS, DSS and OS did not differ significantly between the two groups. Considering the benefits of LSNNS on postoperative quality of life, LSNNS could be recommended as an alternative treatment option for EGC.

Long-term Functional and Patient-reported Outcomes Between Intra-corporeal Delta-shaped Gastroduodenostomy and Gastrojejunostomy After Laparoscopic Distal Gastrectomy.

PURPOSE: This study aimed to compare the long-term functional and patient-reported outcomes between intra-corporeal delta-shaped gastroduodenostomy and gastrojejunostomy after laparoscopic distal gastrectomy for gastric cancer. MATERIALS AND METHODS: We retrospectively reviewed clinicopathological data from 616 patients who had undergone laparoscopic distal gastrectomy for stage I gastric cancer between January 2015 and September 2020. Among them, 232 patients who had undergone delta-shaped anastomosis and another 232 who had undergone Billroth II anastomosis were matched using propensity scores. Confounding variables included age, sex, body mass index, physical status classification, tumor location, and T classification. Postoperative complications, nutritional outcomes, endoscopic findings, and quality of life (QoL) were compared between the 2 groups. RESULTS: No significant differences in postoperative complications or nutritional parameters between the two groups were observed. Annual endoscopic findings revealed more residual food and less bile reflux in the delta group (P<0.001) than in the Billroth II group. Changes of QoL were significantly different regarding emotional function, insomnia, diarrhea, reflux symptoms, and dry mouth (P=0.007, P=0.002, P=0.013, P=0.001, and P=0.03, respectively). Among them, the delta group had worse insomnia, reflux symptoms, and dry mouth within three months postoperatively. CONCLUSIONS: Long-term nutritional outcomes and QoL were comparable between the delta and Billroth II groups. However, more residual food and worse short-term QoL regarding insomnia, reflux symptoms, and dry mouth were observed in the delta group. Longer fasting time before endoscopic evaluation and short-term symptom management would have been helpful for the delta group.

Feature Importance Analysis of a Deep Learning Model for Predicting Late Bladder Toxicity Occurrence in Uterine Cervical Cancer Patients.

(1) In this study, we developed a deep learning (DL) model that can be used to predict late bladder toxicity. (2) We collected data obtained from 281 uterine cervical cancer patients who underwent definitive radiation therapy. The DL model was trained using 16 features, including patient, tumor, treatment, and dose parameters, and its performance was compared with that of a multivariable logistic regression model using the following metrics: accuracy, prediction, recall, F1-score, and area under the receiver operating characteristic curve (AUROC). In addition, permutation feature importance was calculated to interpret the DL model for each feature, and the lightweight DL model was designed to focus on the top five important features. (3) The DL model outperformed the multivariable logistic regression model on our dataset. It achieved an F1-score of 0.76 and an AUROC of 0.81, while the corresponding values for the multivariable logistic regression were 0.14 and 0.43, respectively. The DL model identified the doses for the most exposed 2 cc volume of the bladder (BD2cc) as the most important feature, followed by BD5cc and the ICRU bladder point. In the case of the lightweight DL model, the F-score and AUROC were 0.90 and 0.91, respectively. (4) The DL models exhibited superior performance in predicting late bladder toxicity compared with the statistical method. Through the interpretation of the model, it further emphasized its potential for improving patient outcomes and minimizing treatment-related complications with a high level of reliability.

Predicting factors associated with prolonged intensive care unit stay of patients with COVID-19.

BACKGROUND: Predicting the length of stay (LOS) for coronavirus disease 2019 (COVID-19) patients in the intensive care unit (ICU) is essential for efficient use of ICU resources. We analyzed the clinical characteristics of patients with severe COVID-19 based on their clinical care and determined the predictive factors associated with prolonged LOS. METHODS: We included 96 COVID-19 patients who received oxygen therapy at a high-flow nasal cannula level or above after ICU admission during March 2021 to February 2022. The demographic characteristics at the time of ICU admission and results of severity analysis (Sequential Organ Failure Assessment [SOFA], Acute Physiology and Chronic Health Evaluation [APACHE] II), blood tests, and ICU treatments were analyzed using a logistic regression model. Additionally, blood tests (C-reactive protein, D-dimer, and the PaO2 to FiO2 ratio [P/F ratio]) were performed on days 3 and 5 of ICU admission to identify factors associated with prolonged LOS. RESULTS: Univariable analyses showed statistically significant results for SOFA score at the time of ICU admission, C-reactive protein level, high-dose steroids, mechanical ventilation (MV) care, continuous renal replacement therapy, extracorporeal membrane oxygenation, and prone position. Multivariable analysis showed that MV care and P/F ratio on hospital day 5 were independent factors for prolonged ICU LOS. For D-dimer, no significant variation was observed at admission; however, after days 3 and 5 days of admission, significant between-group variation was detected. CONCLUSIONS: MV care and P/F ratio on hospital day 5 are independent factors that can predict prolonged LOS for COVID-19 patients.

Prolonged patency of fully covered self-expandable metal stents with an externally anchored plastic stent in distal malignant biliary obstruction.

BACKGROUND : Fully covered self-expandable metal stents (FCSEMSs) are widely used for endoscopic treatment of distal malignant biliary obstruction (dMBO). We aimed to assess the efficacy of anchoring an external plastic stent to an FCSEMS in dMBO. METHODS : A multicenter retrospective cohort study was performed in patients with dMBO to compare stent patency between FCSEMSs and FCSEMSs with an externally anchored plastic stent (EPS). For external anchoring, a 7-Fr double-pigtail plastic stent (DPPS) was placed first in the bile duct, then an FCSEMS was deployed side-by-side. RESULTS : Among a total of 185 patients, 65 had an FCSEMS alone and 120 had an FCSEMS with an EPS. The median stent patency was significantly longer in the FCSEMS with an EPS group than in the FCSEMS only group (342 vs. 240 days; P = 0.04). The rate of stent migration was significantly lower in the FCSEMS with an EPS group than in the FCSEMS only group (10.8 % vs. 27.7 %; P = 0.01). There were no significant differences in the rates of stent occlusion and adverse events between the two groups. CONCLUSIONS : A novel and simple technique of anchoring an external plastic stent may decrease the risk of FCSEMS migration and prolong stent patency, without significantly increasing the adverse events rate in dMBO.

Data supporting a saturation mutagenesis assay for Tat-driven transcription with the GigaAssay.

The data in this article are associated with the research paper "GigaAssay - an adaptable high-throughput saturation mutagenesis assay" [1]. The raw data are sequence reads of HIV-1 Tat cDNA amplified from cellular genomic DNA in a new single-pot saturation mutagenesis assay designated the "GigaAssay". A bioinformatic pipeline and parameters used to analyze the data. Raw, processed, analyzed, and filtered data are reported. The data is processed to calculate the Tat-driven transcription activity for cells with each possible single amino acid substitution in Tat. This data can be reused to interpret Tat intermolecular interactions and HIV latency. This is one of the largest and most complete datasets regarding the impact of amino acid substitutions within a single protein on a molecular function.

GigaAssay - An adaptable high-throughput saturation mutagenesis assay platform.

High-throughput assay systems have had a large impact on understanding the mechanisms of basic cell functions. However, high-throughput assays that directly assess molecular functions are limited. Herein, we describe the "GigaAssay", a modular high-throughput one-pot assay system for measuring molecular functions of thousands of genetic variants at once. In this system, each cell was infected with one virus from a library encoding thousands of Tat mutant proteins, with each viral particle encoding a random unique molecular identifier (UMI). We demonstrate proof of concept by measuring transcription of a GFP reporter in an engineered reporter cell line driven by binding of the HIV Tat transcription factor to the HIV long terminal repeat. Infected cells were flow-sorted into 3 bins based on their GFP fluorescence readout. The transcriptional activity of each Tat mutant was calculated from the ratio of signals from each bin. The use of UMIs in the GigaAssay produced a high average accuracy (95%) and positive predictive value (98%) determined by comparison to literature benchmark data, known C-terminal truncations, and blinded independent mutant tests. Including the substitution tolerance with structure/function analysis shows restricted substitution types spatially concentrated in the Cys-rich region. Tat has abundant intragenic epistasis (10%) when single and double mutants are compared.

Laparoscopic Sentinel Node Navigation Surgery for Stomach Preservation in Patients With Early Gastric Cancer: A Randomized Clinical Trial.

PURPOSE: To compare postoperative complications, long-term survival, and quality of life (QOL) after laparoscopic sentinel node navigation surgery (LSNNS) and laparoscopic standard gastrectomy (LSG). METHODS: Five hundred eighty patients with preoperatively diagnosed stage IA gastric adenocarcinoma (≤ 3 cm) were assigned to undergo either LSG or LSNNS. Observers were not blinded to patient grouping. The primary outcome was 3-year disease-free survival (3y-DFS). Secondary outcomes included postoperative complications, QOL, 3-year disease-specific survival (3y-DSS), and 3-year overall survival (3y-OS). RESULTS: In total, 527 patients were included in the modified intention-to-treat analysis population for the primary outcome (LSG, 269; LSNNS, 258). Stomach-preserving surgery was performed in 210 patients (81%) in the LSNNS group. During the median follow-up duration, the 3y-DFS rates in the LSG and LSNNS groups were 95.5% and 91.8%, respectively (difference: 3.7%; 95% CI, -0.6 to 8.1). Three patients with recurrence and five with metachronous gastric cancer in the LSNNS group underwent standard surgery. Two patients with distant metastasis in both groups were treated with palliative chemotherapy. The 3y-DSS and 3y-OS rates in the LSG and LSNNS groups were 99.5% and 99.1% (P = .59) and 99.2% and 97.6% (P = .17), respectively. Postoperative complications occurred in 19.0% of the LSG group and 15.5% of the LSNNS group (P = .294). The LSNNS group showed better physical function (P = .015), less symptoms (P < .001), and improved nutrition than the LSG group. CONCLUSION: LSNNS did not show noninferiority to LSG for 3y-DFS, with a 5% margin. However, the 3y-DSS and 3y-OS were not different after rescue surgery in cases of recurrence/metachronous gastric cancer, and LSNNS had better long-term QOL and nutrition than LSG.

The essential but enigmatic regulatory role of HERVH in pluripotency.

Human specific endogenous retrovirus H (HERVH) is highly expressed in both naive and primed stem cells and is essential for pluripotency. Despite the proven relationship between HERVH expression and pluripotency, there is no single definitive model for the function of HERVH. Instead, several hypotheses of a regulatory function have been put forward including HERVH acting as enhancers, long noncoding RNAs (lncRNAs), and most recently as markers of topologically associating domain (TAD) boundaries. Recently several enhancer-associated lncRNAs have been characterized, which bind to Mediator and are necessary for promoter-enhancer folding interactions. We propose a synergistic model of HERVH function combining relevant findings and discuss the current limitations for its role in regulation, including the lack of evidence for a pluripotency-associated target gene.

Dynamic differential evolution schemes of WRKY transcription factors in domesticated and wild rice.

WRKY transcription factors play key roles in stress responses, growth, and development. We previously reported on the evolution of WRKYs from unicellular green algae to land plants. To address recent evolution events, we studied three domesticated and eight wild species in the genus Oryza, an ideal model due to its long history of domestication, economic importance, and central role as a model system. We have identified prevalence of Group III WRKYs despite differences in breeding of cultivated and wild species. Same groups of WRKY genes tend to cluster together, suggesting recent, multiple duplication events. Duplications followed by divergence may result in neofunctionalizations of co-expressed WRKY genes that finely tune the regulation of target genes in a same metabolic or response pathway. WRKY genes have undergone recent rearrangements to form novel genes. Group Ib WRKYs, unique to AA genome type Oryza species, are derived from Group III genes dated back to 6.76 million years ago. Gene tree reconciliation analysis with the species tree revealed details of duplication and loss events in the 11 genomes. Selection analysis on single copy orthologs reveals the highly conserved nature of the WRKY domain and clusters of fast evolving sites under strong positive selection pressure. Also, the numbers of single copy orthologs under positive or negative selection almost evenly split. Our results provide valuable insights into the preservation and diversification of an important gene family under strong selective pressure for biotechnological improvements of the world's most valued food crop.

Evaluation of Serum Neurofilament Light Chain and Glial Fibrillary Acidic Protein as Screening and Monitoring Biomarkers for Brain Metastases.

We evaluated the potential serum neurofilament light chain (sNfL) and serum glial fibrillary acidic protein (sGFAP) roles in diagnosing and monitoring brain metastases (BMs). We included 70 patients with newly diagnosed BMs, 71 age- and cancer type-matched patients without BMs, and 67 healthy controls (HCs). We compared sNfL and sGFAP levels among the groups using a single-molecule array immunoassay. We prospectively followed 26 patients with BMs every 2-3 months by measuring sNfL and sGFAP levels and performing magnetic resonance imaging (MRI) scans. The sNfL and the sGFAP levels were higher in patients with BMs (medians: sNfL, 63.7 µL; sGFAP, 819.5 pg/µL) than in those without BMs (sNfL, 13.3 µL; sGFAP, 154 pg/µL; p < 0.001) and HCs (sNfL, 12.5 µL; sGFAP, 135 pg/µL; p < 0.001). The sNfL and the sGFAP cutoff levels had a sensitivity and a specificity of 91%. The sGFAP cutoff level had a sensitivity of 91% and a specificity of 97%. The sNfL and the sGFAP levels were related to the BM size but not to the primary cancer type. After BM treatment, sNfL and sGFAP levels decreased with reduced BM lesions on MRI; however, they increased when BMs progressed. sNfL and sGFAP are potential biomarkers for BM screening in cancer patients.

Machine learning approaches for the prediction of bone mineral density by using genomic and phenotypic data of 5130 older men.

The study aimed to utilize machine learning (ML) approaches and genomic data to develop a prediction model for bone mineral density (BMD) and identify the best modeling approach for BMD prediction. The genomic and phenotypic data of Osteoporotic Fractures in Men Study (n = 5130) was analyzed. Genetic risk score (GRS) was calculated from 1103 associated SNPs for each participant after a comprehensive genotype imputation. Data were normalized and divided into a training set (80%) and a validation set (20%) for analysis. Random forest, gradient boosting, neural network, and linear regression were used to develop BMD prediction models separately. Ten-fold cross-validation was used for hyper-parameters optimization. Mean square error and mean absolute error were used to assess model performance. When using GRS and phenotypic covariates as the predictors, all ML models' performance and linear regression in BMD prediction were similar. However, when replacing GRS with the 1103 individual SNPs in the model, ML models performed significantly better than linear regression (with lasso regularization), and the gradient boosting model performed the best. Our study suggested that ML models, especially gradient boosting, can improve BMD prediction in genomic data.

Machine Learning Approaches for Fracture Risk Assessment: A Comparative Analysis of Genomic and Phenotypic Data in 5130 Older Men.

The study aims were to develop fracture prediction models by using machine learning approaches and genomic data, as well as to identify the best modeling approach for fracture prediction. The genomic data of Osteoporotic Fractures in Men, cohort Study (n = 5130), were analyzed. After a comprehensive genotype imputation, genetic risk score (GRS) was calculated from 1103 associated Single Nucleotide Polymorphisms for each participant. Data were normalized and split into a training set (80%) and a validation set (20%) for analysis. Random forest, gradient boosting, neural network, and logistic regression were used to develop prediction models for major osteoporotic fractures separately, with GRS, bone density, and other risk factors as predictors. In model training, the synthetic minority oversampling technique was used to account for low fracture rate, and tenfold cross-validation was employed for hyperparameters optimization. In the testing, the area under curve (AUC) and accuracy were used to assess the model performance. The McNemar test was employed to examine the accuracy difference between models. The results showed that the prediction performance of gradient boosting was the best, with AUC of 0.71 and an accuracy of 0.88, and the GRS ranked as the 7th most important variable in the model. The performance of random forest and neural network were also significantly better than that of logistic regression. This study suggested that improving fracture prediction in older men can be achieved by incorporating genetic profiling and by utilizing the gradient boosting approach. This result should not be extrapolated to women or young individuals.

Acute and Subacute Oral Toxicity of Mumefural, Bioactive Compound Derived from Processed Fruit of Prunus mume Sieb. et Zucc., in ICR Mice.

Mumefural is a bioactive compound derived from the processed fruit of Prunus mume Sieb. et Zucc., a traditional health food; however, its safety has not been evaluated. We investigated the toxicity of mumefural through single and repeated oral administration at doses of 1250, 2500, and 5000 mg/kg in Institute of Cancer Research (ICR) mice. The acute toxicity assessment was not associated with adverse effects or death. Similarly, the subacute (four weeks) toxicity assessment did not reveal any mumefural-associated mortality, abnormal organ damage, or altered clinical signs, body weight, food consumption, or hematological parameters. However, albumin/globulin ratio and chloride ion levels were significantly increased in male mice treated with mumefural at ≥ 2500 mg/kg. Female mice exhibited significantly higher levels of chloride, sodium, and potassium ions, at a dose of 5000 mg/kg. Furthermore, the administration of 2500 and 5000 mg/kg mumefural decreased the absolute weight of spleen in male mice. These findings indicated that the approximate lethal dose of mumefural in ICR mice was > 5000 mg/kg. No significant mumefural toxicity was observed at ≤ 5000 mg/kg. Our findings provide a basis for conducting future detailed studies to evaluate reproductive, neurological, genetic, and chronic toxicity of mumefural.