Category: Statistics

J Agric Food Chem. 2022 Jun 13. doi: 10.1021/acs.jafc.2c01447. Online ahead of print.

ABSTRACT

The price of mammalian milk from different animal species varies greatly due to differences in their yield and nutritional value. Therefore, the authenticity of dairy products has become a hotspot issue in the market due to the replacement or partial admixture of high-cost milk with its low-cost analog. Herein, four common commercial varieties of milk, including goat milk, buffalo milk, Holstein cow milk, and Jersey cow milk, were successfully profiled and differentiated from each other by rapid evaporative ionization mass spectrometry (REIMS) combined with chemometric analysis. This method was developed as a real-time lipid fingerprinting technique. Moreover, the established chemometric algorithms based on multivariate statistical methods mainly involved principal component analysis, orthogonal partial least squares-discriminant analysis, and linear discriminant analysis as the screening and verifying tools to provide insights into the distinctive molecules constituting the four varieties of milk. The ions with m/z 229.1800, 243.1976, 257.2112, 285.2443, 299.2596, 313.2746, 341.3057, 355.2863, 383.3174, 411.3488, 439.3822, 551.5051, 577.5200, 628.5547, 656.5884, 661.5455, 682.6015, and 684.6146 were selected as potential classified markers. The results of the present work suggest that the proposed method could serve as a reference for recognizing dairy fraudulence related to animal species and expand the application field of REIMS technology.

PMID:35696488 | DOI:10.1021/acs.jafc.2c01447

IEEE J Biomed Health Inform. 2022 Jun 13;PP. doi: 10.1109/JBHI.2022.3182869. Online ahead of print.

ABSTRACT

Real-time mental stress monitoring from surgeons and surgical staff in operating rooms may reduce surgical injuries, improve performance and quality of medical care, and accelerate implementation of stress-management strategies. Motivated by the increase in usage of objective and subjective metrics for cognitive monitoring and by the gap in reviews of experimental design setups and data analytics, a systematic review of 71 studies on mental stress and workload measurement in surgical settings, published in 2001-2020, is presented. Almost 61% of selected papers used both objective and subjective measures, followed by 25% that only administered subjective tools – mostly consisting of validated instruments and customized surveys. An overall increase in the total number of publications on intraoperative stress assessment was observed from mid-2010 s along with a momentum in the use of both subjective and real-time objective measures. Cardiac activity, including heart-rate variability metrics, stress hormones, and eye-tracking metrics were the most frequently and electroencephalography (EEG) was the least frequently used objective measures. Around 40% of selected papers collected at least two objective measures, 41% used wearable devices, 23% performed synchronization and annotation, and 76% conducted baseline or multi-point data acquisition. Furthermore, 93% used a variety of statistical techniques, 14% applied regression models, and only one study released a public, anonymized dataset. This review of data modalities, experimental setups, and analysis techniques for intraoperative stress monitoring highlights the initiatives of surgical data science and motivates research on computational techniques for mental and surgical skills assessment and cognition-guided surgery.

PMID:35696473 | DOI:10.1109/JBHI.2022.3182869

IEEE Trans Pattern Anal Mach Intell. 2022 Jun 13;PP. doi: 10.1109/TPAMI.2022.3181472. Online ahead of print.

ABSTRACT

Convolutional neural networks (CNNs) have achieved superior accuracy in many visual-related tasks. However, the inference process through a CNN’s intermediate layers is opaque, making it difficult to interpret such networks or develop trust in their operation. In this paper, we introduce SIGN method for modeling the network’s hidden layer activity using probabilistic models. The activity patterns in layers of interest are modeled as Gaussian mixture models, and transition probabilities between clusters in consecutive modeled layers are estimated to identify paths of inference. For fully connected networks, the entire layer activity is clustered, and the resulting model is a hidden Markov model. For convolutional layers, spatial columns of activity are clustered, and a maximum likelihood model is developed for mining an explanatory inference graph. The graph describes the hierarchy of activity clusters most relevant for network prediction. We show that such inference graphs are useful for understanding the general inference process of a class, as well as explaining the (correct or incorrect) decisions the network makes about specific images. In addition, SIGN provide interesting observations regarding hidden layer activity in general, including the concentration of memorization in a single middle layer in fully connected networks, and a highly local nature of column activities in the top CNN layers.

PMID:35696462 | DOI:10.1109/TPAMI.2022.3181472

PLoS Pathog. 2022 Jun 13;18(6):e1009946. doi: 10.1371/journal.ppat.1009946. Online ahead of print.

ABSTRACT

Venezuelan equine encephalitis virus (VEEV) is a positively-stranded RNA arbovirus of the genus Alphavirus that causes encephalitis in humans. Cynomolgus macaques are a relevant model of the human disease caused by VEEV and are useful in exploring pathogenic mechanisms and the host response to VEEV infection. Macaques were exposed to small-particle aerosols containing virus derived from an infectious clone of VEEV strain INH-9813, a subtype IC strain isolated from a human infection. VEEV-exposed macaques developed a biphasic fever after infection similar to that seen in humans. Maximum temperature deviation correlated with the inhaled dose, but fever duration did not. Neurological signs, suggestive of virus penetration into the central nervous system (CNS), were predominantly seen in the second febrile period. Electroencephalography data indicated a statistically significant decrease in all power bands and circadian index during the second febrile period that returned to normal after fever resolved. Intracranial pressure increased late in the second febrile period. On day 6 post-infection macaques had high levels of MCP-1 and IP-10 chemokines in the CNS, as well as a marked increase of T lymphocytes and activated microglia. More than four weeks after infection, VEEV genomic RNA was found in the brain, cerebrospinal fluid and cervical lymph nodes. Pro-inflammatory cytokines & chemokines, infiltrating leukocytes and pathological changes were seen in the CNS tissues of macaques euthanized at these times. These data are consistent with persistence of virus replication and/or genomic RNA and potentially, inflammatory sequelae in the central nervous system after resolution of acute VEEV disease.

PMID:35696423 | DOI:10.1371/journal.ppat.1009946

PLoS Comput Biol. 2022 Jun 13;18(6):e1010129. doi: 10.1371/journal.pcbi.1010129. Online ahead of print.

ABSTRACT

Single cell Hi-C techniques enable one to study cell to cell variability in chromatin interactions. However, single cell Hi-C (scHi-C) data suffer severely from sparsity, that is, the existence of excess zeros due to insufficient sequencing depth. Complicating the matter further is the fact that not all zeros are created equal: some are due to loci truly not interacting because of the underlying biological mechanism (structural zeros); others are indeed due to insufficient sequencing depth (sampling zeros or dropouts), especially for loci that interact infrequently. Differentiating between structural zeros and dropouts is important since correct inference would improve downstream analyses such as clustering and discovery of subtypes. Nevertheless, distinguishing between these two types of zeros has received little attention in the single cell Hi-C literature, where the issue of sparsity has been addressed mainly as a data quality improvement problem. To fill this gap, in this paper, we propose HiCImpute, a Bayesian hierarchical model that goes beyond data quality improvement by also identifying observed zeros that are in fact structural zeros. HiCImpute takes spatial dependencies of scHi-C 2D data structure into account while also borrowing information from similar single cells and bulk data, when such are available. Through an extensive set of analyses of synthetic and real data, we demonstrate the ability of HiCImpute for identifying structural zeros with high sensitivity, and for accurate imputation of dropout values. Downstream analyses using data improved from HiCImpute yielded much more accurate clustering of cell types compared to using observed data or data improved by several comparison methods. Most significantly, HiCImpute-improved data have led to the identification of subtypes within each of the excitatory neuronal cells of L4 and L5 in the prefrontal cortex.

PMID:35696429 | DOI:10.1371/journal.pcbi.1010129

PLoS Comput Biol. 2022 Jun 13;18(6):e1009598. doi: 10.1371/journal.pcbi.1009598. Online ahead of print.

ABSTRACT

Differential sensitivity analysis is indispensable in fitting parameters, understanding uncertainty, and forecasting the results of both thought and lab experiments. Although there are many methods currently available for performing differential sensitivity analysis of biological models, it can be difficult to determine which method is best suited for a particular model. In this paper, we explain a variety of differential sensitivity methods and assess their value in some typical biological models. First, we explain the mathematical basis for three numerical methods: adjoint sensitivity analysis, complex perturbation sensitivity analysis, and forward mode sensitivity analysis. We then carry out four instructive case studies. (a) The CARRGO model for tumor-immune interaction highlights the additional information that differential sensitivity analysis provides beyond traditional naive sensitivity methods, (b) the deterministic SIR model demonstrates the value of using second-order sensitivity in refining model predictions, (c) the stochastic SIR model shows how differential sensitivity can be attacked in stochastic modeling, and (d) a discrete birth-death-migration model illustrates how the complex perturbation method of differential sensitivity can be generalized to a broader range of biological models. Finally, we compare the speed, accuracy, and ease of use of these methods. We find that forward mode automatic differentiation has the quickest computational time, while the complex perturbation method is the simplest to implement and the most generalizable.

PMID:35696417 | DOI:10.1371/journal.pcbi.1009598

PLoS One. 2022 Jun 13;17(6):e0269706. doi: 10.1371/journal.pone.0269706. eCollection 2022.

ABSTRACT

BACKGROUND: The incidence of childhood ALL in Indonesia is still largely unknown. The widely mentioned statistics from other countries turn out to be only estimated figures. Other data do not specify the types of leukemia and are not specifically focused on children. Therefore, this study aims to pool incidence and mortality statistics from available studies in Indonesia.

METHODS: We searched five different academic databases, including Pubmed, MEDLINE, Cochrane Library, Science Direct, and Google Scholar. Three Indonesian databases, such as the Indonesian Scientific Journal Database (ISJD), Neliti, and Indonesia One Search, were also utilized. Incidence was expressed as per 100,000 children. We used the Newcastle-Ottawa scale (NOS) to assess the quality of cohort studies. The inclusion criteria are cohort studies published in the languages of English or Indonesian. For this analysis, we define children as 0-18 years old.

FINDINGS: The incidence rate for childhood ALL was found to be 4.32 per 100,000 children (95% CI 2.65-5.99) with a prediction interval of 1.98 to 9.42 per 100,000 children. The incidence rate is higher in males, with 2.45 per 100,000 children (95% CI 1.98-2.91) and a prediction interval of 1.90 to 3.16 per 100,000 children. As for females, the incidence rate is 2.05 per 100,000 children (95% CI 1.52-2.77) with a prediction interval of 1.52 to 2.77 per 100,000 children. The mortality of childhood ALL ranges from 0.44 to 5.3 deaths per 100,000 children, while the CFR is 3.58% with varying true effect sizes of 2.84% to 4.52%.

INTERPRETATION: With 79.5 million children living in Indonesia in 2018, this means that there were roughly 3,434 new cases of childhood ALL. An organized effort between multiple sectors is needed to improve the registries of childhood ALL in Indonesia.

PMID:35696384 | DOI:10.1371/journal.pone.0269706

Helicobacter. 2022 Jun 13:e12910. doi: 10.1111/hel.12910. Online ahead of print.

ABSTRACT

Helicobacter pylori infection is strongly associated with chronic gastritis and is probably the main course of chronic inflammation in the gastric mucosa. Gradually, H. pylori gastritis will result in gastric atrophy and intestinal metaplasia. Identifying the relationship between intensity of colonization and activity of gastritis helps the clinician in more effective treatment and post-treatment follow-ups. The aim of our work was to analyze the relationship between the density of H. pylori colonization of the gastric mucosa and the severity of histological parameters of gastritis (inflammation activity, gastric atrophy, and intestinal metaplasia). This was a prospective monocentric study conducted from January 2020 to December 2020, collecting patients naive to any anti-H. pylori treatment and having a chronic H. pylori infection documented by histological examination. Epidemiological, endoscopic, and anathomopathological data were collected. Ninety-seven patients with a mean age of 42.6 years [18-65 years] and a sex ratio of M/F = 0.64 were included. The density of H. pylori colonization was mild (+) in 43.3% of patients, moderate (++) in 47.4% of patients, and significant (+++) in 9.3% of patients. Nearly, ten per cent of patients had no gastritis, 33% had mild gastritis, 50.5% had moderate gastritis, and 6.2% had severe gastritis. Gastric atrophy and intestinal metaplasia were found in 44.3% and 10.3% of our population, respectively. Patients with mild H. pylori colonization rates had the highest level of mild activity (59.5%). There was a statistically significant association between the severity of H. pylori infection and gastritis activity (p < .001). Gastric atrophy was significantly associated with the intensity of H. pylori colonization (p = .049). No significant relationship was found between the intensity of colonization and metaplasia (p = .08). Our study shows that there is a statistically significant association between the density of H. pylori and histopathological findings including gastritis activity and intestinal atrophy.

PMID:35696278 | DOI:10.1111/hel.12910

Int J Gynaecol Obstet. 2022 Jun 13. doi: 10.1002/ijgo.14305. Online ahead of print.

ABSTRACT

OBJECTIVE: The aim of this study is to examine the diagnostic value of ophthalmic arteries Doppler indices in prediction of preeclampsia along with other markers in the third trimester of pregnancy.

METHODS: Normotensive subjects were included during 28-32 weeks’ gestation to undergo uterine and ophthalmic artery Doppler ultrasound. Maternal and fetal characteristics were documented at the visit between the 28-32 weeks of gestation, and the pregnancy associated plasma protein-A (PAPP-A) values in the first trimester were collected to be integrated in a multiparametric prediction model.

RESULTS: Out of 795 included participants, 48 cases progressed to preeclampsia. All assessed ophthalmic Doppler parameters including first and second peak systolic velocities (1^st and 2^nd PSV), second to first PSV Ratio (PR) and pulsatility index (PI)) were statistically different in those developing preeclampsia later on. Average PR (sensitivity: 100%, (95%CI): 0.81-1.00, specificity: 90%, 0.86-0.93) and PI between the eyes, PAPP-A MoM and uterine arteries PI were determined to be the most important predictors of PE which were subsequently integrated in a multiple regression model (sensitivity: 94%, 0.70-1.00, specificity: 93%, 0.89-0.96).

CONCLUSION: This study provided a screening method for individuals at higher risk of progressing towards preeclampsia in the third trimester of pregnancy.

PMID:35696254 | DOI:10.1002/ijgo.14305

Diabetes Care. 2022 Jun 13:dc211789. doi: 10.2337/dc21-1789. Online ahead of print.

ABSTRACT

OBJECTIVE: To identify novel biomarkers of cardiovascular disease (CVD) risk in type 2 diabetes (T2D) via a hypothesis-free global metabolomics study, while taking into account renal function, an important confounder often overlooked in previous metabolomics studies of CVD.

RESEARCH DESIGN AND METHODS: We conducted a global serum metabolomics analysis using the Metabolon platform in a discovery set from the Joslin Kidney Study having a nested case-control design comprising 409 individuals with T2D. Logistic regression was applied to evaluate the association between incident CVD events and each of the 671 metabolites detected by the Metabolon platform, before and after adjustment for renal function and other CVD risk factors. Significant metabolites were followed up with absolute quantification assays in a validation set from the Joslin Heart Study including 599 individuals with T2D with and without clinical evidence of significant coronary heart disease (CHD).

RESULTS: In the discovery set, serum orotidine and 2-piperidinone were significantly associated with increased odds of incident CVD after adjustment for glomerular filtration rate (GFR) (odds ratio [OR] per SD increment 1.94 [95% CI 1.39-2.72], P = 0.0001, and 1.62 [1.26-2.08], P = 0.0001, respectively). Orotidine was also associated with increased odds of CHD in the validation set (OR 1.39 [1.11-1.75]), while 2-piperidinone did not replicate. Furthermore, orotidine, being inversely associated with GFR, mediated 60% of the effects of declining renal function on CVD risk. Addition of orotidine to established clinical predictors improved (P < 0.05) C statistics and discrimination indices for CVD risk (ΔAUC 0.053, rIDI 0.48, NRI 0.42) compared with the clinical predictors alone.

CONCLUSIONS: Through a robust metabolomics approach, with independent validation, we have discovered serum orotidine as a novel biomarker of increased odds of CVD in T2D, independent of renal function. Additionally, orotidine may be a biological mediator of the increased CVD risk associated with poor kidney function and may help improve CVD risk prediction in T2D.

PMID:35696261 | DOI:10.2337/dc21-1789