Design of Machine Learning Algorithms and Internal Validation of a Kidney Risk Prediction Model for Type 2 Diabetes Mellitus.

Objective: This study aimed to explore specific biochemical indicators and construct a risk prediction model for diabetic kidney disease (DKD) in patients with type 2 diabetes (T2D). Methods: This study included 234 T2D patients, of whom 166 had DKD, at the First Hospital of Jilin University from January 2021 to July 2022. Clinical characteristics, such as age, gender, and typical hematological parameters, were collected and used for modeling. Five machine learning algorithms [Extreme Gradient Boosting (XGBoost), Gradient Boosting Machine (GBM), Support Vector Machine (SVM), Logistic Regression (LR), and Random Forest (RF)] were used to identify critical clinical and pathological features and to build a risk prediction model for DKD. Additionally, clinical data from 70 patients (nT2D = 20, nDKD = 50) were collected for external validation from the Third Hospital of Jilin University. Results: The RF algorithm demonstrated the best performance in predicting progression to DKD, identifying five major indicators: estimated glomerular filtration rate (eGFR), glycated albumin (GA), Uric acid, HbA1c, and Zinc (Zn). The prediction model showed sufficient predictive accuracy with area under the curve (AUC) values of 0.960 (95% CI: 0.936-0.984) and 0.9326 (95% CI: 0.8747-0.9885) in the internal validation set and external validation set, respectively. The diagnostic efficacy of the RF model (AUC = 0.960) was significantly higher than each of the five features screened with the highest feature importance in the RF model. Conclusion: The online DKD risk prediction model constructed using the RF algorithm was selected based on its strong performance in the internal validation.

CircPTEN-MT from PTEN regulates mitochondrial energy metabolism.

Phosphatase and tensin homolog (PTEN) is a multifunctional gene that is involved in a variety of physiological and pathological processes. Circular RNAs (circRNAs) are generated from back-splicing events during mRNA processing and participate in cell biological processes through binding to RNAs or proteins. However, PTEN-related circRNAs are largely unknown. Here we report that circPTEN- mitochondria (MT) (hsa_circ_0002934) is a circular RNA encoded by exons 3, 4, and 5 of PTEN and is a critical regulator of mitochondrial energy metabolism. CircPTEN-MT is localized to mitochondria and physically associated with leucine-rich pentatricopeptide repeat-containing protein (LRPPRC), which regulates posttranscriptional gene expression in mitochondria. Knocking down circPTEN-MT reduces the interaction of LRPPRC and steroid receptor RNA activator (SRA) stem-loop interacting RNA binding protein (SLIRP) and inhibits the polyadenylation of mitochondrial mRNA, which decreases the mRNA level of the mitochondrial complex Ι subunit and reduces mitochondrial membrane potential and adenosine triphosphate production. Our data demonstrate that circPTEN-MT is an important regulator of cellular energy metabolism. This study expands our understanding of the role of PTEN, which produces both linear and circular RNAs with different and independent functions.

Inert Group-Containing Electrolyte Additive Enabling Stable Aqueous Zinc-Ion Batteries.

Aqueous zinc ion batteries (AZIBs) are considered promising energy storage devices because of their high theoretical energy density and cost-effectiveness. However, the ongoing side reactions and zinc dendrite growth during cycling limit their practical application. Herein, trisodium methylglycine diacetate (Na3MGDA) additive containing the additional inert group methyl is introduced for Zn anode protection, and the contribution of methyl as an inert group to the Zn anode stability is discussed. Experimental results reveal that the methyl group with various effects enhances the interaction between the polar groups in Na3MGDA and the Zn2+/Zn anode. Thus, the polar carboxylate negative ions in MGDA anions can more easily modify the solvation structure and adsorb on the anode surface in situ to establish a hydrophobic electrical double layer (EDL) layer with steric hindrance effects. Such the EDL layer exhibits a robust selectivity for Zn deposition and a significant inhibition of parasitic reactions. Consequently, the Zn||Zn symmetric battery presents 2375 h at 1 mA cm-2, 1 mAh cm-2, and the Zn||V6O13 full battery provides 91% capacity retention after 1300 cycles at 3 A g-1. This study emphasizes the significant role of inert groups of the additive on the interfacial stability during the plating/stripping of high-performance AZIBs.

Efficacy and safety of 13 surgical techniques for the treatment of complex anal fistula, non-Crohn CAF: a systematic review and network meta-analysis.

BACKGROUND: Considering the difficulty of treating complex anal fistula (CAF), various surgical techniques exist in clinical work. However, none are ideal. Evidence on the efficacy and safety of different surgical treatments is scarce. The authors aimed to compare the outcomes of the 13 surgical techniques and tried to find the best surgical method for treating CAF. MATERIALS AND METHODS: The authors searched worldwide databases, including Pubmed, Embase, Cochrane Library, Web of Science, CNKI, WanFang, VIP, and SinoMed, from inception to March 2023. All randomized controlled trials comparing the outcomes of 13 surgical techniques were included according to the PICO principles. The indicators of the cure rate, the recurrence rate, the complication rate, the operating time, the postoperative pain on day 1 (VAS), and the postoperative incontinence in month 1 (Wexner) were extracted and analyzed using STATA software 15.1, Review Manager 5.4, and GeMTC14.3. RESULTS: Twenty-eight randomized controlled trials with a total of 2274 patients were included in the network meta-analysis. There was no statistically significant difference in the comparison among any surgical interventions in terms of the cure rate ( P >0.05 Table 2) and recurrence rate ( P >0.05 Table 3). However, in terms of complication rate, fistulectomy was lower than FPS (Median: 0.14; 95% CI: 0.02-0.70) or fistulotomy (Median: 0.09; 95% CI: 0.01-0.55), and fistulotomy was lower than EAFR (Median: 0.24; 95% CI: 0.05-0.84), LIFT (Median: 0.17; 95% CI: 0.02-0.66) or LIFT-EAFR (Median: 0.11; 95% CI: 0.01-0.69) ( P >0.05 Table 4). The surface estimated the advantages and disadvantages under the cumulative ranking (SUCRA). The ranking results indicated that fistulectomy might have the lowest complication rate (SUCRA=7.9%). Because the network results of the operating time, the postoperative pain, and the postoperative incontinence contained no closed loops, the results of their probability ranking could only be referenced, demonstrating that fistulectomy might have the shortest operating time (SUCRA=23.4%), video-assisted modified ligation of the intersphincteric fistula tract (VAMLIFT) might have the lowest postoperative pain on day 1 (VAS) (SUCRA=0.4%) and LIFT might have the lowest postoperative incontinence in month 1(Wexner) (SUCRA=16.2%). CONCLUSION: Fistulectomy might have the lowest complication rate, which might be the relatively superior surgical technique for treating CAF.

Evaluation of Trace Elements Levels and Construction of Auxiliary Prediction Model in Patients with Diabetes Ketoacidosis in Type 1 Diabetes.

Background: Trace elements play an important role in reflecting physical metabolic status, but have been rarely evaluated in diabetes ketoacidosis (DKA). Since clinical biochemical parameters are the first-line diagnostic data mastered by clinical doctors and DKA has a rapid progression, it is crucial to fully utilize clinical data and combine innovative parameters to assist in assessing disease progression. The aim of this study was to evaluate the levels of trace elements in DKA patients, followed by construction of predictive models combined with the laboratory parameters. Methods: A total of 96 T1D individuals (48 DKA patients) were collected from the First Hospital of Jilin University. Serum calcium (Ca), magnesium (Mg), zinc (Zn), copper (Cu), iron (Fe) and selenium (Se) were measured by Inductively Coupled Plasma Mass Spectrometry, and the data of biochemical parameters were collected from the laboratory information system. Training and validation sets were used to construct the model and examine the efficiency of the model. The lambda-mu-sigma method was used to evaluate the changes in the model prediction efficiency as the severity of the patient's condition increases. Results: Lower levels of serum Mg, Ca and Zn, but higher levels of serum Fe, Cu and Se were found in DKA patients. Low levels of total protein (TP), Zn and high levels of lipase would be an efficient combination for the prediction of DKA (Area under curves for training set and validation set were 0.867 and 0.961, respectively). The examination test confirmed the clinical applicability of the constructed models. The increasing predictive efficiency of the model was found with NACP. Conclusion: More severe oxidative stress in DKA led to further imbalance of trace elements. The combination of TP, lipase and Zn could predict DKA efficiently, which would benefit the early identification and prevention of DKA to improve prognosis.

Development and comparison of machine learning-based models for predicting heart failure after acute myocardial infarction.

AIMS: Heart failure (HF) is one of the common adverse cardiovascular events after acute myocardial infarction (AMI), but the predictive efficacy of numerous machine learning (ML) built models is unclear. This study aimed to build an optimal model to predict the occurrence of HF in AMI patients by comparing seven ML algorithms. METHODS: Cohort 1 included AMI patients from 2018 to 2019 divided into HF and control groups. All first routine test data of the study subjects were collected as the features to be selected for the model, and seven ML algorithms with screenable features were evaluated. Cohort 2 contains AMI patients from 2020 to 2021 to establish an early warning model with external validation. ROC curve and DCA curve to analyze the diagnostic efficacy and clinical benefit of the model respectively. RESULTS: The best performer among the seven ML algorithms was XgBoost, and the features of XgBoost algorithm for troponin I, triglycerides, urine red blood cell count, γ-glutamyl transpeptidase, glucose, urine specific gravity, prothrombin time, prealbumin, and urea were ranked high in importance. The AUC of the HF-Lab9 prediction model built by the XgBoost algorithm was 0.966 and had good clinical benefits. CONCLUSIONS: This study screened the optimal ML algorithm as XgBoost and developed the model HF-Lab9 will improve the accuracy of clinicians in assessing the occurrence of HF after AMI and provide a reference for the selection of subsequent model-building algorithms.

Machine learning predicts the risk of hemorrhagic transformation of acute cerebral infarction and in-hospital death.

BACKGROUND: The incidence of hemorrhagic transformation (HT) during thrombolysis after acute cerebral infarction (ACI) is very high. We aimed to develop a model to predict the occurrence of HT after ACI and the risk of death after HT. METHODS: Cohort 1 is divided into HT and non-HT groups, to train the model and perform internal validation. All first laboratory test results of study subjects were used as features to be selected for machine learning, and the models built by four machine learning algorithms were compared to screen the best algorithm and model. Following that, the HT group was divided into death and non-death for subgroup analysis. Receiver operating characteristic (ROC) curves etc. to evaluate the model. ACI patients in cohort 2 for external validation. RESULTS: In cohort 1, the HT risk prediction model HT-Lab10 built by the XgBoost algorithm performed the best with AUCROC=0.95 (95% CI, 0.93-0.96). Ten features were included in the model, namely B-type natriuretic peptide precursor, ultrasensitive C-reactive protein, glucose, absolute neutrophil value, myoglobin, uric acid, creatinine, Ca2+, Thrombin time, and carbon dioxide combining power. The model also had the ability to predict death after HT with AUCROC=0.85 (95% CI, 0.78-0.91). The ability of HT-Lab10 to predict the occurrence of HT as well as death after HT was validated in cohort 2. CONCLUSIONS: The model HT-Lab10 built using the XgBoost algorithm showed excellent predictive ability in both the occurrence of HT and the risk of HT death, achieving a model with multiple uses.

Deep Learning Enables Optofluidic Zoom System with Large Zoom Ratio and High Imaging Resolution.

Due to the relatively low optical power of a liquid lens, it is usually difficult to achieve a large zoom ratio and a high-resolution image simultaneously in an optofluidic zoom imaging system. We propose an electronically controlled optofluidic zoom imaging system combined with deep learning, which achieves a large continuous zoom change and a high-resolution image. The zoom system consists of an optofluidic zoom objective and an image-processing module. The proposed zoom system can achieve a large tunable focal length range from 4.0 mm to 31.3 mm. In the focal length range of 9.4 mm to 18.8 mm, the system can dynamically correct the aberrations by six electrowetting liquid lenses to ensure the image quality. In the focal length range of 4.0-9.4 mm and 18.8-31.3 mm, the optical power of a liquid lens is mainly used to enlarge the zoom ratio, and deep learning enables the proposed zoom system with improved image quality. The zoom ratio of the system reaches 7.8×, and the maximum field of view of the system can reach ~29°. The proposed zoom system has potential applications in camera, telescope and so on.

The trace that is valuable: serum copper and copper to zinc ratio for survival prediction in younger patients with newly diagnosed acute myeloid leukaemia.

PURPOSE: No data on predicting the survival of AML patients based on the level of trace elements in the serum have been presented to date. The aims of this prospective cohort study were as follows: (i) to evaluate the serum Cu and Zn levels in people from Northeast China, (ii) to assess the association between the serum Cu level (SCL) and Cu to Zn ratio (SCZR) and clinical and nutrition data, and (iii) to investigate the predictive values of the SCL and SCZR in newly diagnosed de novo AML patients. METHODS: A total of 105 newly diagnosed AML patients and 82 healthy controls were recruited. The serum Cu and Zn levels were determined by inductively coupled plasma spectrometry. The associations of SCL and SCZR with the survival of these AML patients were assessed by Cox proportional hazards models. RESULTS: Both SCL and SCZR were positively related to the blast percentage of bone marrow and C-reactive protein, negatively related to albumin level and CEBPA double mutation and were significantly associated with worse overall survival and disease-free survival. Meanwhile, patients with higher SCL had worse CTCAE levels, and patients with higher SCZR showed less complete remission during the first course of induction chemotherapy. Moreover, higher SCZR was positively associated with ELN risk stratification, and was negatively associated with haemoglobin level and prognostic nutritional index (PNI). CONCLUSION: The SCL and SCZR are associated with long-term survival in patients with newly diagnosed AML undergoing intensive induction and may serve as important predictive biomarkers.

Effects of Tauroursodeoxycholic Acid and 4-Phenylbutyric Acid on Selenium Distribution in Mice Model with Type 1 Diabetes.

The effect of selenium on diabetes is significant. As pharmaceutical chaperones, tauroursodeoxycholic acid (TUDCA) and 4-phenylbutyric acid (4-PBA) can effectively improve the oxidative stress of the endoplasmic reticulum. This study established a mice model with type 1 diabetes (T1D) to evaluate the effects of pharmaceutical chaperones on selenium distribution. Streptozotocin was used to induce Friend virus B-type mice to establish a T1D mice model. Mice were administered with TUDCA or 4-PBA. Selenium levels in different tissues were measured by inductively coupled plasma-mass spectroscopy (ICP-MS). After treatment with TUDCA and 4-PBA, related laboratory findings such as glucose and glycated serum protein were significantly reduced and were closer to normal levels. At 2 weeks, 4-PBA normalized selenium levels in the heart, and 4-PBA and TUDCA maintained the selenium in the liver, kidney, and muscle at normal. At 2 months, 4-PBA and TUDCA maintained the selenium in the heart, liver, and kidney at normal levels. The serum selenium had a positive correlation with zinc and copper in the diabetes group and the control group, while the serum selenium had no significant association with magnesium and calcium at 2 weeks and 2 months. TUDCA and 4-PBA have crucial effects on selenium distribution in diabetic mice, and further research is needed to research their internal mechanisms.

Characterization and analysis of linear epitopes corresponding to SARS-CoV-2 outbreak in Jilin Province, China.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its variants have caused hundreds of thousands of deaths and shown serious social influence worldwide. Jilin Province, China, experienced the first wave of the outbreak from December 2020 to February 2021. Here, we analyzed the genomic characteristics of the SARS-CoV-2 outbreak in Jilin province using a phylogeographic tree and found that clinical isolates belonged to the B.1 lineage, which was considered to be the ancestral lineage. Several dominant SARS-CoV-2 specific linear B cell epitopes that reacted with the convalescent sera were also analysed and identified using a peptide microarray composed of S, M, and E proteins. Moreover, the serum of convalescent patients infected with SARS-CoV-2 showed neutralizing activity against four widely spreading SARS-CoV-2 variants; however, significant differences were observed in neutralizing activities against different SARS-CoV-2 variants. These data provide important information on genomic characteristics, linear epitopes, and neutralizing activity of SARS-CoV-2 outbreak in Jilin Province, China, which may aid in understanding disease patterns and regional aspects of the pandemic.

Combined Diagnosis of SARS-CoV-2: Rapid Antigen Detection as an Adjunct to Nucleic Acid Detection.

Coronavirus disease 2019 is a serious threat to human life, and early diagnosis and screening can help control the COVID-19 pandemic. The high sensitivity of reverse transcriptase-polymerase chain reaction (RT-PCR) assay is the gold standard for the diagnosis of COVID-19, but there are still some false-negative results. Rapid antigen detection (RAD) is recommended by the World Health Organization (WHO) as a screening method for COVID-19. This review analyzed the characteristics of RDT and found that although the overall sensitivity of RAD was not as high as that of RT-PCR, but RAD was more sensitive in COVID-19 patients within 5 days of the onset of symptoms and in COVID-19 patients with Ct ≤ 25. Therefore, RAD can be used as an adjunct to RT-PCR for screening patients with early COVID-19. Finally, this review provides a combined diagnostic protocol for RAD and nucleic acid testing with the aim of providing a feasible approach for COVID-19 screening.

Evaluation among trace elements, clinical parameters and type 1 diabetes according to sex: A new sight of auxiliary prediction in negative insulin auto-antibodies population.

BACKGROUND: Type 1 diabetes (T1D) exhibited sex-specific metabolic status including oxidative stress with dynamic change of trace elements, which emphasized the importance of the evaluation of trace elements according to sex. Besides, the most significant characteristic, insulin auto-antibodies, could not be found in all T1D patients, which needed the auxiliary prediction of clinical parameters. And it would benefit the early detection and treatment if some high-risk groups of T1D could predict and prevent the occurrence of disease through common clinical parameters. Hence, there was an urgent need to construct more effective and scientific statistical prediction models to serve clinic better. This study aimed to evaluate the sex-specific levels of trace elements and the relationship between trace elements and clinical parameters in T1D, and construct sex-specific auxiliary prediction model combined with trace elements and clinical parameters. METHODS: A total of 105 T1D patients with negative insulin auto-antibodies and 105 age/sex-matched healthy individuals were enrolled in First Hospital of Jilin University. Inductively Coupled Plasma Mass Spectrometry was performed for the measurement of calcium (Ca), magnesium (Mg), zinc (Zn), copper (Cu), iron (Fe), selenium (Se) in the serum, and the data of clinical parameters were received from medical record system. The lambda-mu-sigma method was used to evaluate the relationship between abnormal clinical parameters and trace elements. Training set and validation set were divided for the construction of predictable models in males and females: clinical parameters model, trace element model and the combined model (clinical parameters and trace elements). Goodness fit test, decision curve analysis and other related statistical methods were used to perform data analysis. RESULTS: Lower levels of Mg, Ca, Fe in the serum were found in T1D population in females compared with healthy population, while levels of Fe, Zn and Cu of serum in T1D individuals were higher than those of healthy population in males. Levels of serum Mg, Fe and Cu in T1D group were found with significant sex difference for (P < 0.05), and the levels of Fe and Cu in serum of males were higher than those of females, level of serum Mg in males was lower than those of females. Levels of serum Mg and Zn showed fluctuation trend with increased numbers of abnormal clinical parameters (NACP) in males. Serum Zn in females showed consistent elevated trend with NACP; serum Se increased first and then decreased with NACP in males and females. The auxiliary prediction model (Triglyceride, Total protein, serum Mg) was found with the highest predicted efficiency in males (AUC=0.993), while the model in females (Apolipoprotein A, Creatinine, Fe, Se, Zn/Cu ratio) showed the best predicted efficiency (AUC=0.951). The models had passed the verification in validation set, and Chi-square goodness-of-fit test, DCA results both confirmed their satisfactory clinical applicability. CONCLUSION: Sex-specific difference were found in serum Mg, Fe and Cu in T1D. The combination of triglyceride, total protein and serum Mg for males, and apolipoprotein A, creatinine, Fe, Se, Zn/Cu ratio for females could effectively predict T1D in patients with negative anti-bodies, which would provide alarm for the population with high-risk of T1D and serve the T1D prediction in patients with negative anti-bodies.

The efficacy and safety of Chinese herbal medicine for reducing wound complications after anal fistula surgery: A protocol for systematic review and meta-analysis.

BACKGROUND: Anal fistula is one of the most common diseases in anorectal surgery. The wound healing after surgery can affect the prognosis of patients. We conducted a protocol for systematic review and meta-analysis to assess the efficacy and safety of Chinese herbal medicine for reducing wound complications after anal fistula surgery. METHODS: We have prepared this protocol in accordance with the Preferred Reporting Item for Systematic Review and Meta-analysis (PRISMA-P) statement. We will search the following databases: the China National Knowledge Infrastructure, Wanfang Database, Chinese Science and Technology Periodical Database, Chinese Biomedical Literature Database, Pubmed, Embase, Web of Science, and the Cochrane library. Two authors will independently assess the risk of bias of the included studies based on the bias risk assessment tool recommended in the Cochrane "Risk of bias" assessment tool. All calculations are carried out with STATA13.0 software. RESULTS: A synthesis of current evidence of Chinese herbal medicine for wound management after anal fistula surgery will be shown in this protocol. CONCLUSION: This study may provide more convincing evidence to help clinicians make decisions when dealing with anal fistula patients after surgery.

Comparison of the performance of two real-time fluorescent quantitative PCR kits for the detection of SARS-CoV-2 nucleic acid: a study based on large real clinical samples.

BACKGROUND: The global pandemic of coronavirus disease 2019 (COVID-19) has led to the development of multiple detection kits by national manufacturers for severe acute respiratory syndrome coronavirus 2 viral nucleic acid testing. The purpose of this study is to evaluate the performance of different kits (i.e., Maccura kit and Sansure kit) in real clinical work using clinical samples, which will help with the optimization of the test kits. METHOD: During the past three months (March-May 2022), 1399 pharyngeal swabs from suspected COVID-19 patients have been initially screened using the Maccura kit in Jilin, China, and the test results were verified using the Sansure kit. The cycle threshold (Ct) values generated by the two kits were compared at different viral load levels. Correlation and consistency of the Ct values were investigated using Spearman correlation, Deming regression, and Bland-Altman plots. The cut-off Ct values of the Maccura kit were recalculated by referencing the result of the Sansure kit as a standard. Furthermore, another 163 pharyngeal swabs from suspected COVID-19 patients were collected to verify the new cut-off values. RESULTS: As a result of the Maccura kit testing, 1192 positive cases and 207 suspected COVID-19 cases were verified. After re-examination by the Sansure kit, 1118 positive cases were confirmed. The difference between the Ct values provided by the two kits was statistically significant, except for the N gene at high viral load. The Ct values obtained from the two kits presented a linear positive correlation. The Maccura kit used new cut-off Ct values of 35.00 (ORF1ab gene) and 35.07 (N gene). Based on that, the validation pass rate for the new cut-off Ct values was 91.41%. CONCLUSION: Since the Maccura kit is found to have false positives in actual clinical work, recalculation of the cut-off values can reduce this occurrence. In order to improve the accuracy of the testing, laboratories should use two kits for COVID-19 testing, and the adjusting and optimizing of the kits for their situation are needed.

Evaluating the role of SARS-CoV-2 target genes based on two nucleic acid assay kits.

Background: Effective isolation and early treatment of coronavirus disease 2019 (COVID-19) relies on rapid, accurate, and straightforward diagnostic tools. In response to the rapidly increasing number of cases, reverse transcription-quantitative polymerase chain reaction (RT-qPCR) assays for multiple target genes have become widely available in the market. Methods: In total, 236 COVID-19 patients with positive results in both RT-qPCR and rapid antigen diagnosis (Ag-RDT) were enrolled in the study. The cycle threshold (Ct) was compared with different onset times and target genes. Comparison between groups was evaluated with the Kruskal-Wallis test and Dunn test. The correlation between target genes was analyzed by Spearman. Results: In samples of Ct ≤ 21, Ct was different for the nucleocapsid (N), open reading frame 1ab (ORF1ab), and envelope (E) genes (P < 0.05). Mild COVID-19 patients within 7 days of onset accounted for 67.80% of all enrolled patients. At the above stage, all target genes reached the trough of Ct, and N genes showed lower values than the other target genes. The Ct of the ORF1ab and N gene in asymptomatic patients differed from those of mild patients within 7 days and more than 14 days of onset. The kits used in the study showed strong consistency among target genes, with all correlation coefficients >0.870. Conclusion: RT-qPCR confirmed that the N gene performed well in Ct ≤ 21 and samples within 7 days of onset. Ag-RDT was discriminatory for patients within 7 days of onset. This study facilitated early identification and control of COVID-19 prevalence among patients.

A complete blood count-based multivariate model for predicting the recovery of patients with moderate COVID-19: a retrospective study.

Many resource-limited countries need an efficient and convenient method to assess disease progression in patients with coronavirus disease 2019 (COVID-19). This study developed and validated a complete blood count-based multivariate model for predicting the recovery of patients with moderate COVID-19. We collected the clinical data and laboratory test results of 86 patients with moderate COVID-19. These data were categorized into two subgroups depending on the laboratory test time. Univariate logistic regression and covariance diagnosis were used to screen for independent factors, and multifactorial logistic regression was used for model building. Data from 38 patients at another hospital were collected for external verification of the model. Basophils (OR 6.372; 95% CI 3.284-12.363), mean corpuscular volume (OR 1.244; 95% CI 1.088-1.422), red blood cell distribution width (OR 2.585; 95% CI 1.261-5.297), and platelet distribution width (OR 1.559; 95% CI 1.154-2.108) could be combined to predict recovery of patients with moderate COVID-19. The ROC curve showed that the model has good discrimination. The calibration curve showed that the model was well-fitted. The DCA showed that the model is clinically useful. Small increases in the above parameters within the normal range suggest an improvement in patients with moderate COVID-19.

Potential Predictive Value of miR-125b-5p, miR-155-5p and Their Target Genes in the Course of COVID-19.

Purpose: This study aimed to provide new biomarkers for predicting the disease course of COVID-19 by analyzing the dynamic changes of microRNA (miRNA) and its target gene expression in the serum of COVID-19 patients at different stages. Methods: Serum samples were collected from all COVID-19 patients at three time points: the acute stage, the turn-negative stage, and the recovery stage. The expression level of miRNA and the target mRNA was measured by Quantitative Real-Time Polymerase Chain Reaction (RT-qPCR). The classification tree model was established to predict the disease course, and the prediction efficiency of independent variables in the model was analyzed using the receiver operating characteristic (ROC) curve. Results: The expression of miR-125b-5p and miR-155-5p was significantly up-regulated in the acute stage and gradually decreased in the turn-negative and recovery stages. The expression of the target genes CDH5, STAT3, and TRIM32 gradually down-regulated in the acute, turn-negative, and recovery stages. MiR-125b-5p, miR-155-5p, STAT3, and TRIM32 constituted a classification tree model with 100% accuracy of prediction and AUC >0.7 for identification and prediction in all stages. Conclusion: MiR-125b-5p, miR-155-5p, STAT3, and TRIM32 could be useful biomarkers to predict the time nodes of the acute, turn-negative, and recovery stages of COVID-19.