ipl-logo

D-Xylitol Lab Report

855 Words4 Pages

4 Discussion At the beginning of this experiment, simulation and manipulation process are required in order to construct new model of E. coli to be used in this experiment. The new constructed model of E. coli consists of only several pathways which are glycolysis and PPP. Even without glycolysis itself, E. coli still can survive with PPP pathway only. This is as stated by [20], where they mentioned that PPP can be used as a bypass pathway which is the pathway that can be used with additional reactions when the typical glycolysis pathway is not available. Main reason on why PPP has been the only focusing pathway in this experiment is because the pathway that will produce xylitol in E. coli contain in PPP only. As stated by [10, 21], PPP is …show more content…

Since the pathway to produce D-xylose was in PPP which is the main pathway that is needed to produce xylitol and that is why the new constructed model of E. coli focusing only on PPP. After the new model of E. coli has been constructed with several modifications, the predictive analysis by using DNN was conducted in order to predict which gene knockout condition is the best to be applied in the E. coli so that it can produce more xylitol. The experiment has been conducted using several iterations which are 50, 100, 150 and 200 in order to measure and observe the consistency of the predicted values for each of the gene knockout conditions. By the end of this process, DNN method has predicted that with an absence of PGI, TKTa and TKTb from E. coli metabolic pathway, xylitol can be produced at the high level. In contrast, when experiment has been carried out by using DNNDSA, the predicted best gene knockout condition is when only both PGI and TKTa absence from the E. coli …show more content…

As can be seen from the result between DNN and DNNDSA, DNN method can give the output within short time period compared to DNNDSA where it takes longer time to give the output. DNNDSA may require longer time compared to DNN because it is a hybrid method which consists both DNN and DSA. Therefore, the complexity of the algorithm is bigger compared to DNN as well as their search space. It requires longer time to complete the searching process in order to find the best optimal solution because it needs to search for the best solution in both local and global minima space. The search space is double wider compared to DNN where DNN only need to find the optimal solution within local minima search space. This explains why the computational time of DNNDSA is longer compared to DNN

Open Document