Machine Learning Approach Using KPCA-SVMs for Predicting COVID-19

Use of SARS-CoV-2-infected deceased organ donors: Should we always "just say no?"

Even though extreme containment and mitigation strategies were implemented by numerous governments around the world to slow down the spread of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the number of critically ill patients and fatalities keeps rising. This crisis has highlighted the socioeconomic disparities of health care systems within and among countries. As new CoVID policies and responses are implemented to lessen the impact of the virus, it is imperative (1) to consider additional mitigation strategies critical for the development of effective countermeasures, (2) to promote long-term policies and strict regulations of the trade of wildlife and live animal markets, and (3) to advocate for necessary funding and investments in global health, specifically for the prevention of and response to natural and manmade pandemics. This document considers some of these challenges.

SARS-CoV-2's origin should be investigated worldwide for pandemic prevention

If severe acute respiratory syndrome (SARS) recurs, we will need accurate clinical strategies for detecting early cases so that we don't overburden the health care and public health systems with pa...

Peripartum use of Extracorporeal Membrane Oxygenation (ECMO) in a Patient Suffering from COVID-19 Severe Acute Respiratory Distress Syndrome (ARDS): A Case Report

Integrating near-infrared hyperspectral imaging with machine learning and feature selection: Detecting adulteration of extra-virgin olive oil with lower-grade olive oils and hazelnut oil

SARS-CoV-2 and Zoonotic Preparedness: Unknown Knowns?

Gut Involvement by COVID-19

COVID-19 and Gastrointestinal Tract Symptoms: Recognition, Containment, and Learning From the Past

In December 2019, a novel coronavirus was identified in Wuhan City, Hubei Province, China and later the disease was named coronavirus disease 2019 (COVID-19). On March 11, 2020, the World Health Organization (WHO) officially announced that COVID-19 had reached global pandemic status. This article summarized the understanding of the etiology, pathogenesis, epidemiology, clinical characteristics, diagnosis, treatment, rehabilitation, and prevention and control measures of COVID-19 based on the available data and anti-epidemic experience in China.

Transmission of infection from non-isolated patients with COVID-19 to healthcare workers

The outbreak of severe acute respiratory syndrome (SARS) coronavirus 2 (SARS-CoV-2) has resulted in a global pandemic with an exponential growth rate. A recent study found SARS-CoV-2 nucleocapsid protein in kidney tubules of patients infected with SARS-CoV-2 (B. Diao, C.H. Wang, R.S. Wang, Z.Q. Feng, Y.J. Tan, H.M. Wang, et al.: Human kidney is a target for novel severe acute respiratory syndrome coronavirus 2 [SARS-CoV-2] infection [preprint posted online April 10, 2020]. medRxiv doi:10.1101/2020.03.04.20031120), providing direct evidence of kidney susceptibility to SARS-CoV-2. However, the full extent of clinical manifestations with kidney disease and prognoses in patients with the infection remains largely unknown. Because CKD affects 10%–15% of the global population, we urgently need to clarify (1) who is susceptible to kidney damage, (2) what the clinical manifestations are and how they should be diagnosed at an early stage, and (3) how the patients should be monitored during follow-up. The recent development of large-scale "omics" analytic approaches provides a large number of publicly available data sets, such as the spatial characterization of the transcriptome and proteome in different tissues of the human body (1) and expression quantitative trait loci (eQTLs) in kidney compartments (2,3). Human angiotensin-converting enzyme 2 (ACE2) has been identified as the functional receptor of SARS-CoV-2 (4). Thus, genetic analysis of the spatial distribution of human ACE2 expression and its genetic determinants in kidney provides a promising opportunity for predicting kidney involvement during SARS-CoV-2 infection at an early stage. We first determined the gene and protein expression levels of human ACE2 along with its spatial characterization in kidney compartments in The Human Protein Atlas (https://www.proteinatlas.org/) (1). Human ACE2 was highly expressed in multiple organs, including the kidney. This was consistent with the fact that, although SARS-CoV-2 infection primarily manifests as acute respiratory illness, it has also been detected in urine samples. In the kidney, human ACE2 was specifically highly expressed in tubules rather than in glomeruli (Figure 1), suggesting tubular injury as the main consequence of SARS-CoV-2 infection in the kidney. This was validated in a recent study that observed macrophage infiltration and acute tubular damage but no severe glomerular injury (B. Diao, C.H. Wang, R.S. Wang, Z.Q. Feng, Y.J. Tan, H.M. Wang, et al.: Human kidney is a target for novel severe acute respiratory syndrome coronavirus 2 [SARS-CoV-2] infection [preprint posted online April 10, 2020]. medRxivdoi:10.1101/2020.03.04.20031120).Figure 1.: Human ACE2 was highly stained in tubules rather than in glomeruli in human kidney tissue. These results were derived from the Human Protein Atlas database (https://www.proteinatlas.org/). Protein expression in normal kidney tissue from six patients was detected by immunohistochemistry using antibodies HPA000288 (rabbit; Sigma) and CAB026174 (mouse; R&D Systems). The ages of patients range from 16 to 70 years old. As shown, human ACE2 was not detected in glomeruli but was highly stained in tubules.We then searched for the genetic determinants of human ACE2 expression in kidney tubulointerstitial tissue to evaluate the susceptibility to kidney damage in patients infected with SARS-CoV-2. Gillies et al. (2) provided thousands of kidney-specific eQTLs in manually microdissected glomerular (n=136) and tubulointerstitial (n=166) compartments from patients with proteinuric kidney diseases, such as FSGS, minimal change disease, and membranous nephropathy (NephQTL database, http://nephqtl.org/). After searching, 49 variants with P<0.05 were detected; the lowest P value detected was 7.89×10−3, but none of the variants achieved a Bonferroni-corrected threshold P<1.02×10−3. Likewise, there were no significant eQTLs from microdissected tubulointerstitial compartments (n=119) from healthy human kidneys of white subjects undergoing surgical nephrectomy (Human Kidney eQTL Atlas, http://susztaklab.com/eqtl) (3). Therefore, the gene expression of human ACE2 in tubules is less likely to be affected by genetic variants, meaning that the general population could potentially be susceptible to consequences from SARS-CoV-2 infection on the kidney. Finally, we applied our genetic analysis pipeline to other viral infections with well known receptors (e.g., hepatitis B and D viruses). As expected, the cellular receptor of the hepatitis B and D viruses, sodium taurocholate cotransporting polypeptide, was highly specifically expressed in human hepatocytes. The strictly hepatotropic character of hepatitis B and D viruses suggests that SARS-CoV-2 may be nephrotropic. In fact, previous studies have shown a considerable degree of kidney involvement in SARS and Middle East respiratory syndrome (MERS), including AKI and kidney failure (5). Both SARS-CoV RNA and viral particles have been observed in kidney tubules from SARS autopsies (6), indicating direct infection and replication in the kidney. Dipeptidyl peptidase-4, the cellular receptor of MERS-CoV, is also expressed in kidney tubules. There is also evidence supporting kidney infection and induction of apoptosis by MERS-CoV in human ex vivo organ culture (7). The underlying mechanism of kidney involvement in patients infected with SARS-CoV-2 is still unclear, but the existing evidence suggests that the direct infection of SARS-CoV-2 in kidney tubules may play a role. It is still unclear whether the kidney may be a hidden reservoir for SARS-CoV-2 and if the virus will persistently replicate in the kidney. Thus, more attention should be given to the long-term sequelae of SARS-CoV-2 infection and intensive monitoring of kidney function seems necessary. However, it should be noted that, in contrast to the relatively high expression of the sodium taurocholate cotransporting polypeptide receptor of hepatitis B and D viruses in blood cells (normalized expression ranges from 0.1 to 2.7), the normalized expression levels of human ACE2 ranges from 0.0 to 0.3 (https://www.proteinatlas.org/), implying that SARS-CoV-2 has a relatively low viremic rate. It was reported that viral RNA was detectable by real-time PCR of plasma samples in about 15% (two of 13) of patients in the intensive-care unit and 14% (four of 28) of other patients (8). However, the clinical significance of viral RNA in plasma for affecting kidney manifestations and whether the virus itself is also present in plasma must still be determined. Thus, although SARS-CoV-2 showed nephrotropic effects, the rate of kidney injury directly caused by SARS-CoV-2 needs further evaluation. The spatial characteristics of RNA, protein expression of human ACE2, and kidney-specific eQTL analysis indicate that SARS-COV-2 could affect the kidneys of the general population infected with the virus, tubular injury might be the main pathologic manifestation of kidney involvement, and there is a necessity for intensive monitoring of kidney function during follow-up. However, because of a lack of routine screening and monitoring of kidney function, the exact incidence of kidney involvement is still unclear. Current studies have mainly focused on AKI, which is defined by increased serum creatinine or urine output criteria, with an incidence rate varying from 0.3% to 15%. Although, as mentioned above, there might be a relatively lower degree of direct kidney effects of SARS-CoV-2 compared with the effects of hepatitis B and D viruses on the liver. SARS-CoV-2–related cytokine storm and sepsis that lead to tubular and endothelial injury may also be potential pathways of kidney damage (e.g., microthrombi) (9). Thus, to determine the exact incidence rate of kidney injury due to SARS-CoV-2 infection, it is necessary to screen patients who are infected for signs of kidney damage by performing tests commonly used in the clinic, such as measuring serum creatinine or performing dipstick tests for proteinuria and hematuria. Additionally, because human ACE2 is strictly expressed in kidney tubules, the early kidney damage biomarkers such as neutrophil gelatinase-associated lipocalin, kidney injury molecule-1, and IL-18 could also be considered for further clinical evaluation. Furthermore, because accumulated SARS-CoV-2 nucleocapsid protein was detected in kidney tubules from autopsies performed on patients with SARS-CoV-2, patients diagnosed with de novo kidney disease could be tested for virus RNA or viral antigen in urine as a possible surrogate for early diagnosis or for monitoring of disease severity. Finally, it needs to be determined whether patients with preexisting kidney disease are more susceptible to a progressive disease course or a higher risk of acute kidney failure. A global effort to better understand kidney injury due to SARS-CoV-2 is warranted. In conclusion, we provide a genetic pipeline to help explore the roadmap for kidney involvement in SARS-CoV-2 infection. This study highlights the benefits of integrating omics data sets with eQTLs to identify the roles of target genes in disease pathogenesis; this approach will also be applicable for other viruses with well known receptors. Disclosures Dr. H. Zang and Dr. Y. Zhang have nothing to disclose. Funding This study was supported by National Natural Science Foundation of China grant 81800636.

Covid-19, ACE2 and the kidney.

Anemia is one of the common adverse reactions after hip fracture surgery. The traditional method to solve anemia is allogeneic transfusion. However, the transfusion may lead to some complications such as septicemia and fever. So far, few studies have reported roles of machine learning in predicting whether blood transfusion is needed or not after hip fracture surgery. Therefore, the purpose of this study is to develop machine learning models to predict the likelihood of postoperative blood transfusion in patients undergoing hip fracture surgery. This study enrolled 1355 patients who underwent hip fracture surgery at the Affiliated Hospital of Qingdao University from January 2016 to December 2021. Among all patients, 210 cases received postoperative blood transfusion. All patients were randomly divided into a training group and a testing group at a ratio of 7:3. In the training group, univariate and multivariate logistic regression analyses were used to determine independent risk factors for the postoperative transfusion. Then, based on these independent risk factors, tenfold cross-validation method was utilized to develop five machine learning models, including logistic, multilayer perceptron (MLP), extreme gradient boosting (XGBoost), random forest (RF), and support vector machine (SVM). The receiver operating characteristic (ROC) curve, area under ROC curve (AUC), and Matthews correlation coefficient (MCC) were generated to evaluate the performance of the models. Calibration plot and decision curve analysis (DCA) were used to test the performance, stability, and clinical applicability of the models. The models were validated using the testing group; and the ROC curve, MCC, calibration plot, and DCA curves were also generated to validate the performance, stability, and clinical applicability of the models. To further verify the robustness of the model, we randomly grabbed 70% of the samples in the testing set, performed 1000 iterations, and calculated the AUC and confidence interval of the five models. Finally, we used SHapley Additive exPlanations (SHAP) to explain these models. Multivariate logistic regression analysis showed that there were 8 independent risk factors, including age, blood transfusion history, albumin (ALB), globulin (GLO), total bilirubin (TBIL), indirect bilirubin (IBIL), hemoglobin (HB), and blood loss > 200ml. We finally selected five independent risk factors including HB, GLO, age, IBIL, and blood loss > 200ml. Based on these five independent risk factors, we generated six characteristic variables, namely HB, HB × HB, HB × blood loss, GLO × HB, age, age × IBIL, and established five machine learning models using a tenfold cross-validation method. In the training group, the AUC values of logistic, RF, MLP, SVM, and XGB were 0.9320, 0.8911, 0.9327, 0.9225, and 0.8825, respectively, and the average AUC was 0.9122 ± 0.0212. The MCC values were 0.65, 0.77, 0.65, 0.66, and 0.68, respectively, and the calibration plot and DCA performed well. In the testing group the AUC values of logistic, RF, MLP, SVM, and XGB were 0.8483, 0.7978, 0.8576, 0.8598, and 0.8216, respectively. The average AUC was 0.8370 ± 0.0238, and the MCC values were 0.41, 0.35, 0.40, 0.41, and 0.41, respectively. The calibration plot and DCA in the testing group also showed good performance. The AUC values and confidence intervals of the 1000-iteration model were: logistic (AUC, min confidence interval [CI]-max confidence interval [CI] 0.848, 0.804-0.903), RF (AUC, minCI-maxCI 0.797, 0.734-0.857), MLP (AUC, minCI-maxCI 0.858, 0.812-0.902), SVM (AUC, minCI-maxCI 0.859, 0.819-0.910), and XGB (AUC, minCI-maxCI 0.821, 0.764-0.894). The model performed well. Finally, according to SHAP, among all five models, HB played the most important role in model prediction and interpretation. The five models we developed all performed well in predicting the likelihood of blood transfusion after hip fracture surgery. Therefore, we believed that the prediction model based on machine learning had great application prospects in clinical practice, which could help clinicians better predict the risk of blood transfusion after hip fracture surgery.

Recombinant nucleocapsid protein-based IgG enzyme-linked immunosorbent assay for the serological diagnosis of SARS