Confidence-based laboratory test reduction recommendation algorithm.

BACKGROUND: We propose a new deep learning model to identify unnecessary hemoglobin (Hgb) tests for patients admitted to the hospital, which can help reduce health risks and healthcare costs. METHODS: We collected internal patient data from a teaching hospital in Houston and external patient data from the MIMIC III database. The study used a conservative definition of unnecessary laboratory tests, which was defined as stable (i.e., stability) and below the lower normal bound (i.e., normality). Considering that machine learning models may yield less reliable results when trained on noisy inputs containing low-quality information, we estimated prediction confidence to assess the reliability of predicted outcomes. We adopted a "select and predict" design philosophy to maximize prediction performance by selectively considering samples with high prediction confidence for recommendations. Our model accommodated irregularly sampled observational data to make full use of variable correlations (i.e., with other laboratory test values) and temporal dependencies (i.e., previous laboratory tests performed within the same encounter) in selecting candidates for training and prediction. RESULTS: The proposed model demonstrated remarkable Hgb prediction performance, achieving a normality AUC of 95.89% and a Hgb stability AUC of 95.94%, while recommending a reduction of 9.91% of Hgb tests that were deemed unnecessary. Additionally, the model could generalize well to external patients admitted to another hospital. CONCLUSIONS: This study introduces a novel deep learning model with the potential to significantly reduce healthcare costs and improve patient outcomes by identifying unnecessary laboratory tests for hospitalized patients.

Non-invasive arterial blood pressure measurement and SpO2 estimation using PPG signal: a deep learning framework.

BACKGROUND: Monitoring blood pressure and peripheral capillary oxygen saturation plays a crucial role in healthcare management for patients with chronic diseases, especially hypertension and vascular disease. However, current blood pressure measurement methods have intrinsic limitations; for instance, arterial blood pressure is measured by inserting a catheter in the artery causing discomfort and infection. METHOD: Photoplethysmogram (PPG) signals can be collected via non-invasive devices, and therefore have stimulated researchers' interest in exploring blood pressure estimation using machine learning and PPG signals as a non-invasive alternative. In this paper, we propose a Transformer-based deep learning architecture that utilizes PPG signals to conduct a personalized estimation of arterial systolic blood pressure, arterial diastolic blood pressure, and oxygen saturation. RESULTS: The proposed method was evaluated with a subset of 1,732 subjects from the publicly available ICU dataset MIMIC III. The mean absolute error is 2.52 ± 2.43 mmHg for systolic blood pressure, 1.37 ± 1.89 mmHg for diastolic blood pressure, and 0.58 ± 0.79% for oxygen saturation, which satisfies the requirements of the Association of Advancement of Medical Instrumentation standard and achieve grades A for the British Hypertension Society standard. CONCLUSIONS: The results indicate that our model meets clinical standards and could potentially boost the accuracy of blood pressure and oxygen saturation measurement to deliver high-quality healthcare.

Sulphur dioxide and fluoride co-exposure induce incisor hypomineralization and amelogenin upregulation via YAP/RUNX2 signaling pathway.

Sulphur dioxide (SO2) and fluoride are among the most common environmental pollutants affecting human health, and both co-exist in areas predominantly consuming coal. It is vital to analyse the combined toxicity of SO2 and fluoride, and their effects on health and the underlying mechanisms of their co-exposure have not yet been adequately assessed. In the present study, we used ICR mice and LS8 cells to investigate the toxicity of SO2 and fluoride exposure to the enamel, alone or in combination. Factorial design analysis was used to reveal the combined toxicity in vitro and in vivo. Co-exposure to SO2 and fluoride exacerbated enamel injury, resulting in more severe hypomineralization of incisor, and enamel structure disorders in mice, and could induce the accumulation of protein residue in the matrix of the enamel. Amelogenin expression was increased upon exposure to SO2 and fluoride, but enamel matrix proteases were not affected. Consistent with our in vivo results, co-exposure of SO2 and fluoride aggravated amelogenin expression in LS8 cells, and increased the YAP and RUNX2 levels. Co-exposure to SO2 and fluoride resulted in greater toxicity than individual exposure, both in vitro and in vivo, indicating that residents of areas exposed to SO2 and fluoride may have an increased risk of developing enamel damage.

Population stratification enables modeling effects of reopening policies on mortality and hospitalization rates.

OBJECTIVE: Study the impact of local policies on near-future hospitalization and mortality rates. MATERIALS AND METHODS: We introduce a novel risk-stratified SIR-HCD model that introduces new variables to model the dynamics of low-contact (e.g., work from home) and high-contact (e.g., work on-site) subpopulations while sharing parameters to control their respective R0(t) over time. We test our model on data of daily reported hospitalizations and cumulative mortality of COVID-19 in Harris County, Texas, from May 1, 2020, until October 4, 2020, collected from multiple sources (USA FACTS, U.S. Bureau of Labor Statistics, Southeast Texas Regional Advisory Council COVID-19 report, TMC daily news, and Johns Hopkins University county-level mortality reporting). RESULTS: We evaluated our model's forecasting accuracy in Harris County, TX (the most populated county in the Greater Houston area) during Phase-I and Phase-II reopening. Not only does our model outperform other competing models, but it also supports counterfactual analysis to simulate the impact of future policies in a local setting, which is unique among existing approaches. DISCUSSION: Mortality and hospitalization rates are significantly impacted by local quarantine and reopening policies. Existing models do not directly account for the effect of these policies on infection, hospitalization, and death rates in an explicit and explainable manner. Our work is an attempt to improve prediction of these trends by incorporating this information into the model, thus supporting decision-making. CONCLUSION: Our work is a timely effort to attempt to model the dynamics of pandemics under the influence of local policies.

Meta-analysis of genome-wide association studies and functional assays decipher susceptibility genes for gastric cancer in Chinese populations.

OBJECTIVE: Although a subset of genetic loci have been associated with gastric cancer (GC) risk, the underlying mechanisms are largely unknown. We aimed to identify new susceptibility genes and elucidate their mechanisms in GC development. DESIGN: We conducted a meta-analysis of four genome-wide association studies (GWASs) encompassing 3771 cases and 5426 controls. After targeted sequencing and functional annotation, we performed in vitro and in vivo experiments to confirm the functions of genetic variants and candidate genes. Moreover, we selected 33 promising variants for two-stage replication in 7035 cases and 8323 controls from other five studies. RESULTS: The meta-analysis of GWASs identified three loci at 1q22, 5p13.1 and 10q23.33 associated with GC risk at p<5×10-8 and replicated seven known loci at p<0.05. At 5p13.1, the risk rs59133000[C] allele enhanced the binding affinity of NF-κB1 (nuclear factor kappa B subunit 1) to the promoter of PRKAA1, resulting in a reduced promoter activity and lower expression. The knockout of PRKAA1 promoted both GC cell proliferation and xenograft tumour growth in nude mice. At 10q23.33, the rs3781266[C] and rs3740365[T] risk alleles in complete linkage disequilibrium disrupted and created, respectively, the binding motifs of POU2F1 and PAX3, resulting in an increased enhancer activity and expression of NOC3L, while the NOC3L knockdown suppressed GC cell growth. Moreover, two new loci at 3q11.2 (OR=1.21, p=4.56×10-9) and 4q28.1 (OR=1.14, p=3.33×10-11) were associated with GC risk. CONCLUSION: We identified 12 loci to be associated with GC risk in Chinese populations and deciphered the mechanisms of PRKAA1 at 5p13.1 and NOC3L at 10q23.33 in gastric tumourigenesis.

Phosphorylation of Histone H2A at Serine 95: A Plant-Specific Mark Involved in Flowering Time Regulation and H2A.Z Deposition.

Phosphorylation of histone H3 affects transcription, chromatin condensation, and chromosome segregation. However, the role of phosphorylation of histone H2A remains unclear. Here, we found that Arabidopsis thaliana MUT9P-LIKE-KINASE (MLK4) phosphorylates histone H2A on serine 95, a plant-specific modification in the histone core domain. Mutations in MLK4 caused late flowering under long-day conditions but no notable phenotype under short days. MLK4 interacts with CIRCADIAN CLOCK ASSOCIATED1 (CCA1), which allows MLK4 to bind to the GIGANTEA (GI) promoter. CCA1 interacts with YAF9a, a co-subunit of the Swi2/Snf2-related ATPase (SWR1) and NuA4 complexes, which are responsible for incorporating the histone variant H2A.Z into chromatin and histone H4 acetylase activity, respectively. Importantly, loss of MLK4 function led to delayed flowering by decreasing phosphorylation of H2A serine 95, along with attenuated accumulation of H2A.Z and the acetylation of H4 at GI, thus reducing GI expression. Together, our results provide insight into how phosphorylation of H2A serine 95 promotes flowering time and suggest that phosphorylation of H2A serine 95 modulated by MLK4 is required for the regulation of flowering time and is involved in deposition of the histone variant H2A.Z and H4 acetylation in Arabidopsis.

Phosphorylation of SPT5 by CDKD;2 Is Required for VIP5 Recruitment and Normal Flowering in Arabidopsis thaliana.

The elongation factor suppressor of Ty 5 homolog (Spt5) is a regulator of transcription and histone methylation. In humans, phosphorylation of SPT5 by P-TEFb, a protein kinase composed of Cyclin-dependent kinase 9 (CDK9) and cyclin T, interacts with the RNA polymerase II-associated factor1 (PAF1) complex. However, the mechanism of SPT5 phosphorylation is not well understood in plants. Here, we examine the function of SPT5 in Arabidopsis thaliana and find that spt5 mutant flowers early under long-day and short-day conditions. SPT5 interacts with the CDK-activating kinase 4 (CAK4; CDKD;2) and is specifically phosphorylated by CDKD;2 at threonines. The phosphorylated SPT5 binds VERNALIZATION INDEPENDENCE5 (VIP5), a subunit of the PAF1 complex. Genetic analysis showed that VIP5 acts downstream of SPT5 and CDKD;2 Loss of SPT5 or CDKD;2 function results in early flowering because of decreased amounts of FLOWERING LOCUS C (FLC) transcript. Importantly, CDKD;2 and SPT5 are required for the deposition of VIP5 and the enhancement of trimethylation of histone 3 lysine 4 in the chromatin of the FLC locus. Together, our results provide insight into the mechanism by which the Arabidopsis elongation factor SPT5 recruits the PAF1 complex via the posttranslational modification of proteins and suggest that the phosphorylation of SPT5 by CDKD;2 enables it to recruit VIP5 to regulate chromatin and transcription in Arabidopsis.

MC4R deficiency in pigs results in hyperphagia and ultimately hepatic steatosis without high-fat diet.

Melanocortin 4 receptor (MC4R)-deficient mice had been used for several years to study human nonalcoholic steatohepatitis (NASH). However, although liver pathologic and biochemical indicators have been examined, mice models do not always faithfully display the phenotype of the human disease. In this study, we investigated the MC4R knockout phenotype in miniature pigs. We found that pigs lacking MC4R exhibited hyperorexia, insulin resistance, hyperinsulinemia, disordered lipid metabolism and their livers accumulated significant amounts of fat. We have shown that deletion of MC4R results in hyperphagia and increased body fat, ultimately leading to hepatic steatosis without atherogenic diet.

Exome Array Analysis Identifies Variants in SPOCD1 and BTN3A2 That Affect Risk for Gastric Cancer.

BACKGROUND & AIMS: Several genetic variants have been associated with gastric cancer risk, although these account for only a fraction of cases of gastric cancer. We aimed to identify low-frequency and other genetic variants that determine gastric cancer susceptibility. METHODS: We performed exome array analysis of DNA in blood samples from 1113 patients with gastric cancer, collected at hospitals from 2006 to 2010 in China, and 1848 individuals without cancer (controls) undergoing physical examinations. Among 71,290 variants analyzed (including 25,784 common variants), 24 variants were selected and replicated in an analysis of DNA in blood samples from 4687 additional cases of gastric cancer and 5780 controls. We compared expression of candidate genes in tumor vs normal gastric tissues using data from TCGA and performed functional annotation analyses. An immortalized human gastric epithelial cell line (GES1) and 7 human gastric cancer lines were used to express transgenes, knock down gene expression (with small interfering RNAs), disrupt genes (using the CRISPR/Cas9 system), or assess expression of reporter constructs. We measured cell proliferation, colony formation, invasion, and migration, and assessed growth of xenograft tumors in nude mice. RESULTS: A low-frequency missense variant rs112754928 in the SPOC domain containing 1 gene (SPOCD1; encoding p.Arg71Trp), at 1p35.2, was reproducibly associated with reduced risk of gastric cancer (odds ratio, 0.56; P = 3.48 × 10-8). SPOCD1 was overexpressed in gastric tumors, and knockout of SPOCD1 reduced gastric cancer cell proliferation, invasive activity, and migration, as well as growth of xenograft tumors in nude mice. We also associated the variant rs1679709 at 6p22.1 with reduced risk for gastric cancer (odds ratio, 0.80; P = 1.17 × 10-13). The protective allele rs1679709-A correlated with the surrounding haplotype rs2799077-T-rs2799079-C, which reduced the enhancer activity of this site to decrease expression of the butyrophilin subfamily 3 member A2 gene (BTN3A2). BTN3A2 is overexpressed in gastric tumors, and deletion of BTN3A2 inhibited proliferation, migration, and invasion of gastric cancer cells. CONCLUSIONS: We have associated variants at 1p35.2 and 6p22.1 with gastric cancer risk, indicating a role for SPOCD1 and BTN3A2 in gastric carcinogenesis.

Genetic variants at 9p21.3 are associated with risk of esophageal squamous cell carcinoma in a Chinese population.

Genome-wide association studies have linked genetic variants at 9p21.3 to the risk of multiple cancers. However, the roles of genetic variants at 9p21.3 in esophageal squamous cell carcinoma (ESCC) development are largely unknown. We evaluated the genetic variants at 9p21.3 reported in cancer genome-wide association studies with a case-control study including 2139 ESCC cases and 2273 controls in a Chinese population, and measured the mRNA expression levels of MTAP, CDKN2A, CDKN2B, and CDKN2B-AS1 in paired ESCC tumor and adjacent normal tissues. We found that the G allele of rs7023329 was significantly associated with a decreased risk of ESCC with a per-allele odds ratio of 0.84 (95% confidence interval, 0.77-0.91; P = 2.95 × 10-5 ). The rs7023329-G allele was related to a high expression of MTAP (P = 0.020). The rs1679013-C allele was independently associated with an increased risk of ESCC with a per-allele odds ratio of 1.12 (95% confidence interval, 1.01-1.24; P = 0.039). We also found that the carriers of the risk allele rs1679013-C had lower expression of CDKN2B than non-carriers (P = 0.035). CDKN2B was also significantly downregulated in ESCC tumor tissues compared with adjacent normal tissues (P = 3.50×10-5 ). Therefore, our findings indicate that genetic variants at 9p21.3 may modulate the expression of MTAP and CDKN2B and contribute to ESCC susceptibility. This may further advance our understanding of the 9p21.3 locus in cancer development.