Explore novel machine learning approaches to outperform conventional methods for portable DNA barcoding
The rights of consumers and genuine food processors in terms of food adulteration and fraudulent or deceptive practices are protected by law. Loss of trust in food products and food labels at societal level can cause reputational damage to companies that eventually negatively affects the competitiveness of the whole sector. High value and complex foods with multiple ingredients such as fish stew and other meals are especially susceptible to mislabelling as the consumer cannot easily identify the fish species used. NGS (Next Generation Sequencing) surpasses other DNA methods but has a relative high cost and slow turnaround time. A new, rapid, and portable DNA sequencing technology (MinION) has been recently introduced that promises to tackle the problems, but it is so new and not tested in the detection of adulteration of fish. This concept is currently explored in-house using 50-100 DNA sequences of four barcodes and performing species-level identification using only distance- and tree-based method against the specie targets. The obtained results are comparable but generally lower than the reference method (RT-PCR). However, according to the literature machine learning (ML) approaches outperform the conventional method (distance- and tree-based methods) for DNA barcoding (He et al., 2019). Therefore, there is a clear opportunity to apply several cutting edge ML approaches, from other fields such as Computer Science, and achieve significantly better results.
Start Date: 1 October 2022
Duration: 3 years
How to apply: Applications must be submitted via: https://dap.qub.ac.uk/portal/user/u_login.php
Skills/experience required: We seek a student with a science background such as Biology and with basic coding skills using open source languages such as R or Python. An interest in bioinformatics will be positively viewed.
Note: This project is in competition for DfE funding with a number of other projects. A selection process will determine the strongest candidates across the range of projects, who may then be offered funding for their chosen project.