GMO Detection with Nanobiosensing System Integration of Artificial Intelligence

Abstract

Genetically modified organisms (GMOs) and their products have been in the food and feed sectors for decades. There are almost 32 crops approved in 44 countries. Every country has different legislation regarding the threshold values and allowed GM events. In the case of Turkey, there are not any GMOs that have been given approval for food use except for three microbial food enzymes. In order to determine GMO levels in the specified limits, rapid and sensitive GMO detection is required. The aim of this study is to propose a new method that makes critical decisions regarding GMO and to develop an amplification-free, DNA-based nanobiosensor that will allow rapid, qualitative determination of the Cry1Ac gene in soybean event MON87701. The new methodology with gold nanoparticles (AuNPs) offers an excellent platform based on the genomic DNA (gDNA) sequence of interest’s hybridization with a complementary sequence. gDNA isolated from the Certified Reference Materials (CRM) was prepared with high purity. The probes used were the exact complementary of the gene of interest. Optimal conditions for the GMO nanobiosensor have been determined and with three batches of citrate reduction synthesized AuNPs. Firstly, heat treatment is applied to unamplified gDNA and the complementary probe to allow hybridization, later addition of AuNPs. Detection was accomplished through aggregation of AuNPs which was associated with color changes of the reaction after addition of NaCl. The aggregation levels were evaluated using UV–vis absorption or by visual observation immediately. A correlation was obtained between the GMO level according to the Cry1Ac gene with a colorimetric change of red to purple. The detection limit is as low as nanomolar, and the detection time estimated as 10 min. The method proposed here has the potential to be an alternative method available in food. Finally, a successful prediction analytics model has been developed from the obtained data set with the machine learning algorithm, support vector machine, that will automatically classify and predict GM levels from different batches of AuNPs. Further models are integrated into a user-friendly website.