An application of raman spectroscopy in combination with machine learning to determine gastric cancer spectroscopy marker.

BACKGROUND AND OBJECTIVE: Globally, gastric carcinoma (Gca) ranks fifth in terms of incidence and third in terms of mortality. Higher serum tumor markers (TMs) than those from healthy individuals, led to TMs clinical application as diagnostic biomarkers for Gca. Actually, there is no accurate blood test to diagnose Gca. METHODS: Raman spectroscopy is applied as an efficient, credible, minimally invasive technique to evaluate the serum TMs levels in blood samples. After curative gastrectomy, serum TMs levels are important in predicting the recurrence of gastric cancer, which must be detected early. The experimentally assesed TMs levels using Raman measurements and ELISA test were used to develop a prediction model based on machine learning techniques. A total of 70 participants diagnosed with gastric cancer after surgery (n = 26) and healthy (n = 44) were comrpised in this study. RESULTS: In the Raman spectra of gastric cancer patients, an additional peak at 1182 cm-1 was observed and, the Raman intensity of amide III, II, I, and CH2 proteins as well as lipids functional group was higher. Furthermore, Principal Component Analysis (PCA) showed, that it is possible to distinguish between the control and Gca groups using the Raman range between 800 and 1800 cm-1, as well as between 2700 and 3000 cm-1. The analysis of Raman spectra dynamics in gastric cancer and healthy patients showed, that the vibrations at 1302 and 1306 cm-1 were characteristic for cancer patients. In addition, the selected machine learning methods showed classification accuracy of more than 95%, while obtaining an AUROC of 0.98. Such results were obtained using Deep Neural Networks and the XGBoost algorithm. CONCLUSIONS: The obtained results suggest, that Raman shifts at 1302 and 1306 cm-1 could be spectroscopic markers of gastric cancer.

Six potential biomarkers for bladder cancer: key proteins in cell-cycle division and apoptosis pathways.

BACKGROUND: The bladder cancer (BC) pathology is caused by both exogenous environmental and endogenous molecular factors. Several genes have been implicated, but the molecular pathogenesis of BC and its subtypes remains debatable. The bioinformatic analysis evaluates high numbers of proteins in a single study, increasing the opportunity to identify possible biomarkers for disorders. METHODS: The aim of this study is to identify biomarkers for the identification of BC using several bioinformatic analytical tools and methods. BC and normal samples were compared for each probeset with T test in GSE13507 and GSE37817 datasets, and statistical probesets were verified with GSE52519 and E-MTAB-1940 datasets. Differential gene expression, hierarchical clustering, gene ontology enrichment analysis, and heuristic online phenotype prediction algorithm methods were utilized. Statistically significant proteins were assessed in the Human Protein Atlas database. GSE13507 (6271 probesets) and GSE37817 (3267 probesets) data were significant after the extraction of probesets without gene annotation information. Common probesets in both datasets (2888) were further narrowed by analyzing the first 100 upregulated and downregulated probesets in BC samples. RESULTS: Among the total 400 probesets, 68 were significant for both datasets with similar fold-change values (Pearson r: 0.995). Protein-protein interaction networks demonstrated strong interactions between CCNB1, BUB1B, and AURKB. The HPA database revealed similar protein expression levels for CKAP2L, AURKB, APIP, and LGALS3 both for BC and control samples. CONCLUSION: This study disclosed six candidate biomarkers for the early diagnosis of BC. It is suggested that these candidate proteins be investigated in a wet lab to identify their functions in BC pathology and possible treatment approaches.

Prediction of Alzheimer's Disease by a Novel Image-Based Representation of Gene Expression.

Early intervention can delay the progress of Alzheimer's Disease (AD), but currently, there are no effective prediction tools. The goal of this study is to generate a reliable artificial intelligence (AI) model capable of detecting the high risk of AD, based on gene expression arrays from blood samples. To that end, a novel image-formation method is proposed to transform single-dimension gene expressions into a discriminative 2-dimensional (2D) image to use convolutional neural networks (CNNs) for classification. Three publicly available datasets were pooled, and a total of 11,618 common genes' expression values were obtained. The genes were then categorized for their discriminating power using the Fisher distance (AD vs. control (CTL)) and mapped to a 2D image by linear discriminant analysis (LDA). Then, a six-layer CNN model with 292,493 parameters were used for classification. An accuracy of 0.842 and an area under curve (AUC) of 0.875 were achieved for the AD vs. CTL classification. The proposed method obtained higher accuracy and AUC compared with other reported methods. The conversion to 2D in CNN offers a unique advantage for improving accuracy and can be easily transferred to the clinic to drastically improve AD (or any disease) early detection.

Defining Molecular Treatment Targets for Bladder Pain Syndrome/Interstitial Cystitis: Uncovering Adhesion Molecules.

Bladder pain syndrome/interstitial cystitis (BPS/IC) is a debilitating pain syndrome of unknown etiology that predominantly affects females. Clinically, BPS/IC presents in a wide spectrum where all patients report severe bladder pain together with one or more urinary tract symptoms. On bladder examination, some have normal-appearing bladders on cystoscopy, whereas others may have severely inflamed bladder walls with easily bleeding areas (glomerulations) and ulcerations (Hunner's lesion). Thus, the reported prevalence of BPS/IC is also highly variable, between 0.06% and 30%. Nevertheless, it is rightly defined as a rare disease (ORPHA:37202). The aetiopathogenesis of BPS/IC remains largely unknown. Current treatment is mainly symptomatic and palliative, which certainly adds to the suffering of patients. BPS/IC is known to have a genetic component. However, the genes responsible are not defined yet. In addition to traditional genetic approaches, novel research methodologies involving bioinformatics are evaluated to elucidate the genetic basis of BPS/IC. This article aims to review the current evidence on the genetic basis of BPS/IC to determine the most promising targets for possible novel treatments.

Myophosphorylase (PYGM) mutations determined by next generation sequencing in a cohort from Turkey with McArdle disease.

This study aimed to identify PYGM mutations in patients with McArdle disease from Turkey by next generation sequencing (NGS). Genomic DNA was extracted from the blood of the McArdle patients (n = 67) and unrelated healthy volunteers (n = 53). The PYGM gene was sequenced with NGS and the observed mutations were validated by direct Sanger sequencing. A diagnostic algorithm was developed for patients with suspected McArdle disease. A total of 16 deleterious PYGM mutations were identified, of which 5 were novel, including 1 splice-site donor, 1 frame-shift, and 3 non-synonymous variants. The p.Met1Val (27-patients/11-families) was the most common PYGM mutation, followed by p.Arg576* (6/4), c.1827+7A>G (5/4), c.772+2_3delTG (5/3), p.Phe710del (4/2), p.Lys754Asnfs (2/1), and p.Arg50* (1/1). A molecular diagnostic flowchart is proposed for the McArdle patients in Turkey, covering the 6 most common PYGM mutations found in Turkey as well as the most common mutation in Europe. The diagnostic algorithm may alleviate the need for muscle biopsies in 77.6% of future patients. A prevalence of any of the mutations to a geographical region in Turkey was not identified. Furthermore, the NGS approach to sequence the entire PYGM gene was successful in detecting a common missense mutation and discovering novel mutations in this population study.

Genetic variants in the tumor necrosis factor-related apoptosis-inducing ligand and death receptor genes contribute to susceptibility to bladder cancer.

AIM: The aim of this study was to evaluate the role of polymorphisms of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and death receptor (DR4) genes in bladder cancer susceptibility in a Turkish population. MATERIALS AND METHODS: The study group included 91 bladder cancer patients, while the control group comprised 139 individuals with no evidence of malignancy. Gene polymorphisms of TRAIL C1595T (rs1131580) and DR4 C626G (rs4871857) were genotyped by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. RESULTS: The frequency of the TRAIL 1595 TT genotype was significantly lower in patients with bladder cancer compared to controls (p<0.001; odds ratios [OR]=0.143; 95% confidence interval [CI]=0.045-0.454). A significantly increased risk for developing bladder cancer was found for the group bearing a C allele for TRAIL C1595T polymorphism (p<0.001; OR=1.256; 95% CI=1.138-1.386). The observed genotype and allele frequencies of DR4 626 C/G in all groups were in agreement with the Hardy-Weinberg equilibrium (p=0.540). However, the frequency of DR4 GG genotype was found to be 2.1-fold increased in the bladder cancer patients with high-grade tumor, when compared to those having low-grade tumor (p=0.036). Additionally, combined genotype analysis showed that the frequency of TRAILCT-DR4GG was significantly higher in patients with bladder cancer in comparison with those of controls (p=0.037; OR=2.240; 95% CI=1.138-1.386). CONCLUSIONS: Our study provides new evidence that TRAIL 1595 C allele may be used as a low-penetrant risk factor for bladder cancer development in a Turkish population. Otherwise, gene-gene interaction analysis revealed that the DR4GG genotype may have a predominant effect on the increased risk of bladder cancer over the TRAIL CT genotype.

The effects of PON1 gene Q192R variant on the development of uterine leiomyoma in Turkish patients.

AIM: This study aimed to analyze the relation between uterine leiomyoma (ULM) patients and p.Q192R polymorphism. MATERIALS AND METHODS: ULM patients (n=76) and healthy women (n=103) were recruited from the Yeditepe University, Department of Gynecology and Obstetrics. The genotype and allele distribution of p.Q192R was analyzed by polymerase chain reaction and restriction fragment length polymorphism methods. Genotype and allele frequencies between study groups were calculated by the chi-square (χ(2)) and Fischer's exact test. RESULTS: The frequency of the B allele was lower in patients (p<0.001) and the AB genotype showed a decreased risk for ULM development (p<0.001). The variation was unrelated to ULM size and number. There was no significant difference between p.Q192R genotype frequencies and fibroid size and number. CONCLUSION: The heterogeneous AB genotype of PON1 p.Q192R variation could be recognized as a low-risk parameter for the development of ULM in Turkish women.