mRBioM: An Algorithm for the Identification of Potential mRNA Biomarkers From Complete Transcriptomic Profiles of Gastric Adenocarcinoma.

PURPOSE: In this work, an algorithm named mRBioM was developed for the identification of potential mRNA biomarkers (PmBs) from complete transcriptomic RNA profiles of gastric adenocarcinoma (GA). METHODS: mRBioM initially extracts differentially expressed (DE) RNAs (mRNAs, miRNAs, and lncRNAs). Next, mRBioM calculates the total information amount of each DE mRNA based on the coexpression network, including three types of RNAs and the protein-protein interaction network encoded by DE mRNAs. Finally, PmBs were identified according to the variation trend of total information amount of all DE mRNAs. Four PmB-based classifiers without learning and with learning were designed to discriminate the sample types to confirm the reliability of PmBs identified by mRBioM. PmB-based survival analysis was performed. Finally, three other cancer datasets were used to confirm the generalization ability of mRBioM. RESULTS: mRBioM identified 55 PmBs (41 upregulated and 14 downregulated) related to GA. The list included thirteen PmBs that have been verified as biomarkers or potential therapeutic targets of gastric cancer, and some PmBs were newly identified. Most PmBs were primarily enriched in the pathways closely related to the occurrence and development of gastric cancer. Cancer-related factors without learning achieved sensitivity, specificity, and accuracy of 0.90, 1, and 0.90, respectively, in the classification of the GA and control samples. Average accuracy, sensitivity, and specificity of the three classifiers with machine learning ranged within 0.94-0.98, 0.94-0.97, and 0.97-1, respectively. The prognostic risk score model constructed by 4 PmBs was able to correctly and significantly (∗∗∗ p < 0.001) classify 269 GA patients into the high-risk (n = 134) and low-risk (n = 135) groups. GA equivalent classification performance was achieved using the complete transcriptomic RNA profiles of colon adenocarcinoma, lung adenocarcinoma, and hepatocellular carcinoma using PmBs identified by mRBioM. CONCLUSIONS: GA-related PmBs have high specificity and sensitivity and strong prognostic risk prediction. MRBioM has also good generalization. These PmBs may have good application prospects for early diagnosis of GA and may help to elucidate the mechanism governing the occurrence and development of GA. Additionally, mRBioM is expected to be applied for the identification of other cancer-related biomarkers.

Automatic classification of esophageal disease in gastroscopic images using an efficient channel attention deep dense convolutional neural network.

The accurate diagnosis of various esophageal diseases at different stages is crucial for providing precision therapy planning and improving 5-year survival rate of esophageal cancer patients. Automatic classification of various esophageal diseases in gastroscopic images can assist doctors to improve the diagnosis efficiency and accuracy. The existing deep learning-based classification method can only classify very few categories of esophageal diseases at the same time. Hence, we proposed a novel efficient channel attention deep dense convolutional neural network (ECA-DDCNN), which can classify the esophageal gastroscopic images into four main categories including normal esophagus (NE), precancerous esophageal diseases (PEDs), early esophageal cancer (EEC) and advanced esophageal cancer (AEC), covering six common sub-categories of esophageal diseases and one normal esophagus (seven sub-categories). In total, 20,965 gastroscopic images were collected from 4,077 patients and used to train and test our proposed method. Extensive experiments results have demonstrated convincingly that our proposed ECA-DDCNN outperforms the other state-of-art methods. The classification accuracy (Acc) of our method is 90.63% and the averaged area under curve (AUC) is 0.9877. Compared with other state-of-art methods, our method shows better performance in the classification of various esophageal disease. Particularly for these esophageal diseases with similar mucosal features, our method also achieves higher true positive (TP) rates. In conclusion, our proposed classification method has confirmed its potential ability in a wide variety of esophageal disease diagnosis.

Mechanism of pretreatment using magnetic poly(glycidyl methacrylate) resin in an ultrafiltration membrane system used in algae-rich water treatment.

Ultrafiltration (UF) membrane fouling brought by algae-rich water controlling has been the research focus in recent years. The pretreatment of magnetic poly(glycidyl methacrylate) (m-PGMA) for sedimental tank effluent was investigated as well as its performance in combined UF processes. The optimal dose of m-PGMA was found to be 5 mL/L, which can bring a significant improvement to the removal efficiency of natural organic matter. With regards to membrane fouling, the use of m-PGMA also resulted in lowered irreversible and reversible membrane resistances in comparison with results obtained when operating without m-PGMA. In addition, four classic filtration models were introduced to analyse the fouling mechanisms. The proportion of standard blocking of pores has been weakened in the mechanism of membrane fouling when the pretreatment of m-PGMA exists. A very loose cake layer and relieved pore blockage were observed by scanning electron microscopy during m-PGMA/UF process.

Preparation and characterization of γ-AlOOH @CS magnetic nanoparticle as a novel adsorbent for removing fluoride from drinking water.

For this study, a novel adsorbent of γ-AlOOH @CS (pseudoboehmite and chitosan shell) magnetic nanoparticles (ACMN) with magnetic separation capabilities was developed to remove fluoride from drinking water. The adsorbent was first characterized, and then its performance in removing fluoride was evaluated. Kinetic data demonstrated rapid fluoride adsorption with more than 80% fluoride adsorption within the initial 20 min and equilibrium reached in 60 min. Based on the results of kinetic and isotherm models, the fluoride adsorption process on the ACMN's surface was a monolayer adsorption on a homogeneous surface. Thermodynamic parameters presented that the adsorption process is spontaneous and endothermic in nature. The mechanism for the adsorption involved electrostatic interaction and hydrogen bonding. Moreover, the calculated adsorption capacity of the ACMN for fluoride using the Langmuir model was 67.5 mg/g (20°C, pH=7.0±0.1), higher than other fluoride removal adsorbents. This nanoadsorbent performed well over a pH range of 4-10. The study found that PO4(3-) was the co-existing anion most able to hinder the nanoparticle's fluoride adsorption, followed by NO3(-) then Cl(-). Experimental results suggest that ACMN is a promising adsorbent for treating fluoride-contaminated water.

Flocculation of algal cells by amphoteric chitosan-based flocculant.

A kind of amphoteric chitosan-based flocculant (quaternized carboxymethyl chitosan, denoted as QCMC) has been prepared. QCMC presented significant improvement of water solubility in the whole pH range. The effects of pH, dosage, temperature and original turbidity of algal water on the flocculation performance were investigated. The optimal dosages of QCMC at pH 5, 9 and 12 with original turbidity of 20NTU at 20°C were 0.1, 0.6 and 2.0mg/L, respectively, which were much less than that of chitosan, PAM, Al2(SO4)3 and FeCl3. The floc properties during grow, breakage and regrow period were also evaluated at different pH values in terms of floc size, strength and density. It was demonstrated that QCMC produced larger, stronger and denser flocs than Al2(SO4)3. There is every indication that QCMC is more suitable for algal harvesting than other traditional coagulants or flocculants.