My summer of research has come to an end, but I believe my career as a research scientist is just beginning. This research project has opened my eyes to the ins and outs of a working lab and the fluid nature of research since the facts uncovered alter the course of your work nearly every day.

My research was based on the amphibian disease chytridiomycosis and how certain alleles of the major histocompatibility complex (MHC) class II beta gene confer resistance to chytridiomycosis. Individuals with the resistant allele are more likely to survive the infection, reproduce, and their offspring will most likely have the same resistant allele, ensuring survival. During my first week of research, I had an idea of what I would like to research but little knowledge on the technique to carry out my plans. My supervisor, Professor Aoife McLysaght, aided me in finding the correct tools to use to complete my research. Instead of jumping right into sequence analysis, I was advised to begin a Python programming course. At the time, this felt like an idle task since I was eager to complete my sequence alignments. I later relied on programming and regular expressions to automate tasks, giving me time to focus on other aspects of my work. I completed tutorials provided in the Python for biologists’ course during week one and immediately put these learned skills into use, which was highly rewarding.

I relied on the Genbank sequence database and BLAST (Basic Local Alignment Search Tool) to acquire the sequences to conduct my alignment. This is where I first encountered some unexpected challenges. I used the MHC class II beta-protein sequence from the amphibian Bufo gargarizans to compare against the Genbank database using tBLASTn (comparing the protein sequence from B. gargarizans to the nucleotide sequences available in Genbank) and obtained over 100 results from different amphibian species. To reduce the likelihood of false positives, I recorded results related to the amphibian order Anura (frogs) until the first instance of another class, such as Mammalia. Following this methodology, after the first appearance of Homo sapiens, results for amphibians with less similarity and statistical significance appeared. The evolutionary lineage suggests that species with a closer relation, such as amphibian species, should have a more closely related sequence to each other than to humans. This stopped my work in its tracks. I had to research the evolutionary origins of the major histocompatibility complex (MHC) class II gene to try and come up with a response to this unexpected result, whether it was due to software error or an interesting detail in the evolution of the MHC.

I used Macse (a multiple sequence alignment program) to align the BLAST results based on triplet codons and created a phylogenetic tree of the MHC class II beta gene from humans to amphibians to assess its evolution. Based on the phylogenetic tree produced from the BLAST search using Bufo gargarizans as the reference sequence, the alignment was led astray based on the reduced length of the partial sequence. I revised the BLAST search using a different species with a longer sequence available, Rana yavapaeinsis, and obtained 524 correct sequences from the order Anura with no unexpected results. This troubleshooting did not occur as succinctly as I have described, with many cycles of trial and error occurring to find the correct species to use in my search and assess the quality of the alignments and phylogenetic trees produced by Macse. From taking my time learning Python, I had a newfound sense of importance when it came to acing the basics as they are the foundation on which the rest of the research is built. The week I spent trawling through the literature, comparing phylogenetic trees, and running Macse with different parameters felt like a valuable lesson in sorting issues when they arise, as it would be near impossible to alter results without starting over.

Figure 1. Sequence alignment viewed using Aliview. Accession number from Genbank of sequence (left) and single-letter amino acid code of sequence (right).

I was introduced to Aliview, increasing my repertoire of bioinformatic tools, to visualise the alignment using colours corresponding to specific amino acids (image above). My focus was on amino acid positions 37, 56, 57, and 60, which are responsible for the shape of the P9 binding pocket for antigens in the MHC. Again, my Python programming skills came in handy as I wrote a short script to help count the number of specific amino acids at a site of interest whilst conducting my statistical analysis. It was a huge reward to see that most sequences contained the resistant amino acid sequences at these sites, meaning many frog species are potentially protected from severe infection. Although my research has been completed, I am creating polished graphs and phylogenetic trees to better visualise the results for my research poster.

My research introduced me to new technical skills and software knowledge which will most definitely aid me in my human genetics degree. The support of the lab group was one of the best rewards since I managed to get a taste of lab experience as an undergraduate and participate in lab meetings and presentations where my communications skills grew in strength. Keeping a blog on the scholar’s network was an enjoyable reflection that I kept throughout my research and helped me connect with others interested in STEM from the different universities involved in the Laidlaw programme.

Taking part in research that could potentially improve conservation techniques used to preserve amphibian wildlife was an honour. I would not change the experience, setbacks included since they were valuable in improving my resilience surrounding unexpected challenges, the overarching theme during my time as a researcher, and I gained beneficial skills to have in the scientific arena.