"The project had to considerably change due to the following reasons. In our preliminary lab work with the sewage sample that we were provided by Dr Orsini, from University of Birmingham (UoB), we observed that we could not get as much biomass as expected. Thus, we would have needed to arrange a second visit to the sewage water cleaning site. That also meant that Dr Zepeda would have needed to try different extraction and isolation techniques in the lab. Unfortunately, all those activities had to be halted due to Covid-19. Notably, considering her asthmatic health record, it was not possible for Dr Zepeda to attend to the lab to keep receiving her laboratory training, even after the lockdown restriction measures were relaxed. Thus, Dr Zepeda modified the deliverables of the Fellowship into a more computational project.
Explanations on the modified project:
Due to the need to work from home given the lockdown restrictions and the subsequent consequences of covid-19, the Fellow started working on a purely computational project using ML and NLP techniques (for text and data mining) to find the factors driving the worldwide spread of AMR using available literature. A publication is currently being drafted, and it is at a state close to submission. The Fellow plans to submit by mid-November to the journal PLoS Computational Biology.
This project follows the innovative approach of using NLP techniques to i) develop a pipeline for the text topic classification of scientific articles, in order to obtain those AMR reports with epidemiological impact, and ii) perform feature importance and model explanatory analyses to identify the features driving the publication of reports of AMR infections by the different countries.
To perform the originally planned project, the Fellow took a training on Data Science. This training was useful for her to instead develop this alternative project during the lockdown. In regards to transfer of knowledge, the Fellow has contributed to the development of other projects in the van Schaik group, by providing useful input and feedback to her colleagues during the weekly meetings that Prof van Schaik holds with his group. Due to her data science training, she also established connections with Dr Johnson, from the School of Mathematics, UoB, who invited her to deliver a talk to the PhD math students as part of their seminar series ""Applied Mathematics"".
Also, with the help of Prof. van Schaik, she established contact with researchers from the School of Computer Science, UoB, who agreed to be her collaborators for a Fellowship application she drafted and submitted to the Royal Society (University Research Fellowship).
The Fellow was involved in performing most of the analyses of the project and in establishing collaboration with Dr Moradigaravand, Center for Computational Biology, UoB, co-author in this paper.
In more general terms, the Fellow was smoothly integrated into the host group and received significant support for the development of her research and training. Dr Zepeda held bi-weekly one-to-one meetings with her supervisor to discuss the results of her project, and she managed the financial aspects of her Fellowship with the support of the administrative staff of IMI UoB."
