To obtain a screening tool for colorectal cancer (CRC) based on gut microbiota, we seek here to identify an optimal classifier for CRC detection as well as a novel nonlinear feature selection method for determining the most discriminative microbial species. In this study, the intestinal microflora in feces of 141 patients were modeled using general regression neural networks (GRNNs) combined with the proposed feature selection method. The proposed model led to slightly higher accuracy (AUC=0.911) than previous studies (AUC<0.87). The results show that the Clostridium scindens and Bifidobacterium angulatum are indicators of healthy gut flora and CRC happens to reduce these bacterial species. In addition, Fusobacterium gonidiaformans was found to be closely correlated with the CRC. The occurrence of colorectal adenoma was not sufficiently discriminatory based on fecal microbiota implicating that the change of colonic flora happens in the advanced phase of CRC development rather than initial adenoma. Integrating the proposed model with fecal occult blood test (FOBT), the CRC detection accuracy remained nearly unchanged (AUC=0.915). The performance of proposed method is validated using independent cohorts from America and Austria. Our results suggest that proposed feature selection method combined with GRNN is potentially an accurate method for CRC detection.

No abstract

No abstract

To develop an anticancer drug, the mathematical models are nowadays indispensable because of complex immunological mechanisms defying with high experimentation costs as well as a large number of parameters. Based on immunological theories and vision of experimentation data, a simple and sufficient compartment model is designed that can accurately interpret and predict the effects of dendritic cell (DC)-based immunotherapy in accordance with experimentation data. The model includes effector cells, regulatory T cells, helper T cells, and DCs. A new key feature is the inclusion of immunotherapy with DCs matured with different materials. All the parameters of the model have been optimally obtained by fitting the experimental data using genetic algorithm. The proposed model has been used to predict a near-optimal pattern that minimizes tumor size after vaccination. This pattern has been validated by carrying out the associated in-vivo experimentation. The model recommends maturation materials and doses that activate a small amount of Treg in the early days and a large Th1/Treg ratio in the next days. The performance of the model compared with the previous study was shown to be superior, both qualitatively and quantitatively.

Previous studies have demonstrated that maturation of dendritic cells (DCs) by pathogenic components through pathogen-associated molecular patterns (PAMPs) such as Listeria monocytogenes lysate (LML) or CpG DNA can improve cancer vaccination in experimental models. In this study, a mathematical model based on an artificial neural network (ANN) was used to predict several patterns and dosage of matured DC administration for improved vaccination. The ANN model predicted that repeated co-injection of tumor antigen (TA)-loaded DCs matured with CpG (CpG-DC) and LML (List-DC) results in improved antitumor immune response as well as a reduction of immunosuppression in the tumor microenvironment. In the present study, we evaluated the ANN prediction accuracy about DC-based cancer vaccines pattern in the treatment of Wehi164 fibrosarcoma cancer-bearing mice. Our results showed that the administration of the DC vaccine according to ANN predicted pattern, leads to a decrease in the rate of tumor growth and size and augments CTL effector function. Furthermore, gene expression analysis confirmed an augmented immune response in the tumor microenvironment. Experimentations justified the validity of the ANN model forecast in the tumor growth and novel optimal dosage that led to more effective treatment.