Explainable Machine Learning for COVID-19 Pneumonia Classification With Texture-Based Features Extraction in Chest Radiography

Both reverse transcription-PCR (RT-PCR) and chest X-rays are used for the diagnosis of the coronavirus disease-2019 (COVID-19). However, COVID-19 pneumonia does not have a defined set of radiological findings. Our work aims to investigate radiomic features and classification models to differentiate chest X-ray images of COVID-19-based pneumonia and other types of lung patterns. The goal is to provide grounds for understanding the distinctive COVID-19 radiographic texture features using supervised ensemble machine learning methods based on trees through the interpretable Shapley Additive Explanations (SHAP) approach. We use 2,611 COVID-19 chest X-ray images and 2,611 non-COVID-19 chest X-rays. After segmenting the lung in three zones and laterally, a histogram normalization is applied, and radiomic features are extracted. SHAP recursive feature elimination with cross-validation is used to select features. Hyperparameter optimization of XGBoost and Random Forest ensemble tree models is applied using random search. The best classification model was XGBoost, with an accuracy of 0.82 and a sensitivity of 0.82. The explainable model showed the importance of the middle left and superior right lung zones in classifying COVID-19 pneumonia from other lung patterns.

Relationship Between the Pyroptosis Pathway and Epilepsy: A Bioinformatic Analysis

PurposeThis study aimed to analyse the correlation between the pyroptosis pathway and epilepsy using bioinformatics analysis technology. We analyzed the expression of gasdermin D (GSDMD) and gasdermin E (GSDME), the key molecules of pyroptosis, in kainic acid-induced epileptic mice.MethodsWeighted gene co-expression network analysis (WGCNA) was used to construct a signed co-expression network from expression data to screen gene sets closely related to epilepsy. The correlation between the module and epilepsy was verified through module conservative analysis, gene ontology (GO) annotation analysis, and correlation analysis with known epilepsy genes. We obtained currently recognized pyroptosis-related molecules through literature review, and correlation analysis was used to evaluate their correlation with epilepsy. Differentially expressed gene (DEG) analysis was used to analyse expression changes of pyroptosis-related molecules at the transcriptome level, compared to the sham group. We subsequently established a kainic acid-induced status epilepticus (SE) model in mice and validated the mRNA and protein expression of GSDMD and GSDME, the key molecules of pyroptosis, by quantitative reverse transcription PCR (qRT-PCR) and western blotting (WB).ResultsUsing WGCNA, module conservative analysis, and correlation analysis with known epilepsy genes, we screened out a module (a gene set of interest) closely related to epilepsy that was prominently enriched in immune and inflammatory-related biological processes. Correlation analysis results suggest that pyroptosis-related molecules are closely related to this module, but have no obvious correlation with others. DEG analysis of molecules associated with pyroptosis suggests that most of the pyroptosis-related molecules had significantly increased expression after SE, such as IL1b, Casp1, Casp4, Pycard, Gsdmd, Nlrp3, Aim2, Mefv, Tlr2, Tlr3, and Tlr4. qRT-PCR and WB analysis confirmed that the mRNA and protein levels of GSDMD in the mouse hippocampus were significantly upregulated after SE. The mRNA expression of GSDME was not different between the epilepsy group and sham group. However, the WB results showed that the expression of full-length GSDME was decreased and GSDME-N-terminus were significantly increased after SE.ConclusionsOur study highlights that the pyroptosis pathway may be closely related to epilepsy. GSDMD and GSDME, the key executive molecules of pyroptosis, will help to understand the pathogenesis of epilepsy and aid in discovering new targets for anti-epileptic drug treatments.

Potential HNF1A-AS1/miR-214/INHBA signal regulation mechanism in colorectal cancer based on bioinformatics analysis and validation

Preceding studies have identified that noncoding RNA plays a significant role in the occurrence and development of tumors. Colorectal cancer (CRC) has attracted increasing attention due to its high incidence and mortality rate. Based on Cancer Genome Atlas (TCGA) database analysis, it was found that compared with normal tissues, HNF1A-AS1 and INHBA were highly expressed in CRC tissues; miR-214 was relatively low expressed, and it is predicted to specifically target the3' untranslated region (3'UTR region) of INHBA. Besides, the result was consistent with the quantitative reverse transcription PCR (RT-qPCR) verification results of 17 CRC cases and adjacent tissues collected clinically. Western Blot (WB) manifested that INHBA protein was highly expressed in CRC tissues, which was consistent with the results of CRC cell lines (HT29, SW480). Immunohistochemical (IHC) staining demonstrated that INHBA protein was brownish yellow, overwhelming majority of INHBA protein were located in the cytoplasm, and expression level was significantly higher than that in adjacent tissues. Based on previous studies, the biological process of INHBA-mediated TGF-β/Smad signaling pathway inducing the proliferation and invasion of CRC has been partially confirmed, but the upstream signaling molecules and mechanisms which regulating INHBA remain unclear. Herein, benefiting from bioinformatics, preliminary experimental results and previous research, they provide basis for the follow-up study on the regulation of HNF1A-AS1/miR-214/INHBA signal axis in CRC.

SARS-CoV-2 N Gene G29195T Point Mutation May Affect Diagnostic Reverse Transcription-PCR Detection

Accurate diagnostic detection of SARS-CoV-2 currently depends on the large-scale deployment of RT-PCR assays. SARS-CoV-2 RT-PCR assays target predetermined regions in the viral genomes by complementary binding of primers and probes to nucleic acid sequences in the clinical samples.

Transcriptional and Physiological Analyses to Assess the Effects of a Novel Biostimulant in Tomato

This work aimed to study the effects in tomato (Solanum lycopersicum L.) of foliar applications of a novel calcium-based biostimulant (SOB01) using an omics approach involving transcriptomics and physiological profiling. A calcium-chloride fertilizer (SOB02) was used as a product reference standard. Plants were grown under well-watered (WW) and water stress (WS) conditions in a growth chamber. We firstly compared the transcriptome profile of treated and untreated tomato plants using the software RStudio. Totally, 968 and 1,657 differentially expressed genes (DEGs) (adj-p-value < 0.1 and |log2(fold change)| ≥ 1) were identified after SOB01 and SOB02 leaf treatments, respectively. Expression patterns of 9 DEGs involved in nutrient metabolism and osmotic stress tolerance were validated by real-time quantitative reverse transcription PCR (RT-qPCR) analysis. Principal component analysis (PCA) on RT-qPCR results highlighted that the gene expression profiles after SOB01 treatment in different water regimes were clustering together, suggesting that the expression pattern of the analyzed genes in well water and water stress plants was similar in the presence of SOB01 treatment. Physiological analyses demonstrated that the biostimulant application increased the photosynthetic rate and the chlorophyll content under water deficiency compared to the standard fertilizer and led to a higher yield in terms of fruit dry matter and a reduction in the number of cracked fruits. In conclusion, transcriptome and physiological profiling provided comprehensive information on the biostimulant effects highlighting that SOB01 applications improved the ability of the tomato plants to mitigate the negative effects of water stress.

Targeted gene suppression through double-stranded RNA application using easy-to-use methods in Arabidopsis thaliana

RNA interference (RNAi) is an RNA-dependent gene silencing process that is regulated by the interaction between the RNA-induced silencing complex (RISC) and double-stranded RNA (dsRNA). Exogenous dsRNAs are imported directly into the cytoplasm, where they are cleaved by Dicer into short dsRNA fragments of 20–25 base pairs. These short dsRNA fragments, called small interfering RNAs (siRNAs) have sequence-specific interaction with target genes. The guide strand, onto which siRNAs are incorporated in the RISC interacts with the target mRNA sequence, thereby inducing cleavage and degradation of target messenger RNAs (mRNAs) by ribonucleases. Recent studies have shown that plant dsRNA treatments can induce RNAi. However, the dsRNA application methods and delivery systems involved have not been well examined. In this study, dsRNA was introduced to Arabidopsis thaliana by two methods: dipping and spray. We synthesized two dsRNAs designed to target mRNAs encoding enhanced green fluorescent protein (EGFP). After applying dsRNAs that target EGFP, we found an obvious reduction in GFP expression. This was determined using fluorescence microscopy and quantitative reverse transcription PCR to assess the mRNA levels of the auxin-sensitive reporter DR5-EGFP Arabidopsis thaliana. Our data revealed that applying target gene-specific exogenous dsRNAs can induce suppression of target genes of interest whether the dipping or spray method is used. This study therefore provides a foundation for understanding how to apply and deliver dsRNAs in plants.

Identification of Hotspot Mutations in the N Gene of SARS-CoV-2 in Russian Clinical Samples That May Affect the Detection by Reverse Transcription-PCR

Sensitive and reliable diagnostic test systems based on real-time PCR are of great importance in the fight against the ongoing SARS-CoV-2 pandemic. The genetic variability of the SARS-CoV-2 virus leads to the accumulation of mutations, some of which may affect the sensitivity of modern PCR assays. The aim of this study was to search in Russian clinical samples for new mutations in SARS-CoV-2 gene N that can affect the detection by RT-PCR. In this study, the polymorphisms in the regions of the target gene N causing failed or poor detection of the target N in the RT-PCR assay on 12 selected samples were detected. Sequencing the entire N and E genes in these samples along with other 195 samples that were positive for both target regions was performed. Here, we identified a number of nonsynonymous mutations and one novel deletion in the N gene that affected the ability to detect a target in the N gene as well a few mutations in the E gene of SARS-CoV-2 that did not affect detection. Sequencing revealed that majority of the mutations in the N gene were located in the variable region between positions 193 and 235 aa, inside and nearby the phosphorylated serine-rich region of the protein N. This study highlights the importance of the further characterization of the genetic variability and evolution of gene N, the most common target for detecting SARS-CoV-2. The use of at least two targets for detecting SARS-CoV-2, including one for the E gene, will be necessary for reliable diagnostics.

Detection of fecal coliforms and SARS-CoV-2 RNA in sewage and recreational waters in the Ecuadorian Coast: a call for improving water quality regulation

Wastewater surveillance represents an alternative approach for the diagnosis and early detection of infectious agents of public health importance. This study aimed to evaluate SARS-CoV-2 and other quality markers in oxidation lagoons, estuarine areas and seawater at Guayas and Santa Elena in Ecuador. Sample collections were conducted twice at 42 coastal sites and 2 oxidation lagoons during dry and rainy seasons (2020-2021). Physico-chemical and microbiological parameters were evaluated to determine organic pollution. Quantitative reverse transcription PCR was conducted to detect SARS-CoV-2. Results showed high levels of Escherichia coli and low dissolved oxygen concentrations. SARS-CoV-2 was detected in sea-waters and estuaries with salinity levels between 34.2-36.4 PSU and 28.8 degrees celsius -31.3 degrees celsius. High amounts of fecal coliforms were detected and correlated with the SARS-CoV-2 shedding. We recommend to decentralized autonomous governments in developing countries such as Ecuador to implement corrective actions and establish medium-term mechanisms to minimize a potential contamination route.

Developing an Amplification Refractory Mutation System-Quantitative Reverse Transcription-PCR Assay for Rapid and Sensitive Screening of SARS-CoV-2 Variants of Concern

The current stage of the pandemic, led by SARS-CoV-2 variants of concern (VOCs), underscores the necessity to develop a cost-effective and rapid molecular diagnosis assay to differentiate the VOCs. In this study, over 1 million SARS-CoV-2 genomic sequences of high quality from GISAID were analyzed and a network of the common mutations of the lineages was constructed.

Detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its first variants in fourplex real-time quantitative reverse transcription-PCR assays

The early diagnosis of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections is required to identify and isolate contagious patients to prevent further transmission of SARS-CoV-2. In this study, we present a multitarget real-time TaqMan reverse transcription PCR (rRT-PCR) assay for the quantitative detection of SARS-CoV-2 and some of its circulating variants harboring mutations that give the virus a selective advantage. Seven different primer-probe sets that included probes containing locked nucleic acid (LNA) nucleotides were designed to amplify specific wild-type and mutant sequences in Orf1ab, Envelope (E), Spike (S), and Nucleocapsid (N) genes. Furthermore, a newly developed primer-probe set targeted human β2-microglobulin (B2M) as a highly sensitive internal control for RT efficacy. All singleplex and fourplex assays detected £ 14 copies/reaction of quantified synthetic RNA transcripts, with a linear amplification range of nine logarithmic orders. Primer-probe sets for detection of SARS-CoV-2 exhibited no false-positive amplifications with other common respiratory pathogens, including human coronaviruses NL63, 229E, OC43, and HKU-1. Fourplex assays were evaluated using 160 clinical samples positive for SARS-CoV-2. Results showed that SARS-CoV-2 viral RNA was detected in all samples, including viral strains harboring mutations in the Spike coding sequence that became dominant in the pandemic. Given the emergence of SARS-CoV-2 variants and their rapid spread in some populations, fourplex rRT-PCR assay containing four primer-probe sets represents a reliable approach to allow quicker detection of circulating relevant variants in a single reaction.