Comparative Analysis of Classifiers for Prediction of Epileptic Seizures

Epilepsy is a neurological disease in which people suffer from seizure attack and lose the normal function of brain. Almost 50 million people have epilepsy in the world due to which it has become the most common neurological disease. Early prediction of epilepsy helps patients to avoid epilepsy and live normal life. Many studies have been conducted for the early prediction of epilepsy. However, selection of the most appropriate classifier has always been a question that needs to be resolved. In this study, we are using six classifiers of machine learning which are KNN, Naïve Bayes, Linear Classification Model, Discriminant Analysis Model, Support Vector Machine and Decision Tree, to find the best classifier for the prediction of epileptic seizures, in term of accuracy. Dataset from “Kaggle” was used. Preprocessing and cross-validation of the data was carried out for training and testing of classifiers. The results depict that Naive Bayes classifier has a better average accuracy of 95.739% as compared to other classifiers. The future work of this study is to implement the suggested model in real time, so that the workload of medical members could be reduced.

Characterization of Indigenous Bacterial Strains Involved in the Production of Endo-Polygalacturonases

The current study was carried out to isolate indigenous bacteria for the production of endo-PG. Bacteria were isolated, purified and screened initially on the basis of morphology, pigment production, margins etc. Ten isolates were selected after the initial screening, which were identified from their 16S rRNA gene sequence. The ten isolates were named IEB-1 to IEB-10. DNA was isolated from the bacteria and 16S rRNA gene was amplified by PCR using universal primers like 27F and 1492R, which were modified for cloning in pTXB1. The PCR product was cloned in pTXB1 and the plasmids were sent for sequencing from University of Rochester Center of Genomics. The 16S rRNA gene sequence was blasted on NCBI nucleotide database. The isolates were identified as Bacillus subtilis, Bacillus pumilius, Exiguobacterium acetylicum, Bacillus pumilius, Staphylococcus hominis, Bacillus licheniformis, Serratia rubidaea respectively from IEB-1 to IEB-9. Nucleotide blast of 16S rRNA gene sequence of IEB-10 showed less than 95% resemblances with many of the species. The selected ten bacterial isolates were then screened for the production of the endopolygalacturonase through different steps. Three isolates showed endo-PG activity. Two isolates were identified as Bacillus pumilus and one was identified as Bacillus licheniformis. The later one has already been reported to be endo-PG positive, but Bacillus pumilus has only been reported as pectinase positive but not as endo-PG positive. Gene sequence of endo-PG from B. pumilus has also not been reported earlier. Primers were designed on the basis of the glycoside hydrolase gene from B. pumilus. The PCR product of glycoside hydrolase was cloned in the double digested pTXB1 which was transformed in E. coliBL21competent cells. E. coli colonies containing the cloned “glycoside hydrolase” gene gave a positive pectinase plate screening assay. B. pumilus was used for the production and characterization of endo-PG. Citrus peals, reddish peals and apple peals were used as the only source of nutrients during liquid state fermentation for B. pumilus for the production of endo-PG to make its production cheaper. Response surface methodology (RSM) under central composite design (CCD) was used for the optimization of various parameters, for the production of the endo-PG, including temperature, pH, inoculum size, inoculum age, fermentation period and substrate water ratio. Three dimensional graphs and contour plots were prepared using JMP-12 software. The optimized results showed the best production of the endo-PG at 39 oC, pH of 7.03, 52 hours of fermentation period, 3.16 % substrate water ratio, 6.5mL of inoculum size having 16 hours of inoculum age. Different carbon, nitrogen sources, surfactants and mediators were used to check their effect on the endo-PG production. All of the carbon sources and some of the nitrogen sources gave negative effect on the production of endo-PG. Ca+2 and Zn+2 showed very negative effect on the endo-PG activity, while Mg+2 had not particular effect at its low concentrations, but higher concentrations of Mg+2 also gave negative effect on the endo-PG activity. The enzyme showed maximum activity at pH 7 and 40 oC. A concentration of 70% was found to be the optimum for the partial purification of the endo-PG. The dialyzed partially purified extract was applied to column chromatography for the further purification. The elution showing a positive endo-PG assay was run on the SDS-PAGE to check the purity and molecular weight of the protein. The stained SDS-PAGE gel showed a single band at 45kDa approximately.