Author: Juliana Renales | Major: Biomedical Engineering | Semester: Spring 2023
Over the course of 2023 spring semester, I worked with my mentor, Dr. Chris Nelson from Biomedical Engineering, to establish an analysis pipeline to map gene expression in muscle injuries. Spatial transcriptomics is a relatively new and exciting method that allows us to collect data on gene expression levels and where this activity is located within the tissue sample. This can be extremely helpful when looking at things such as wounds, as we can analyze cell function in the context of what is happening in the wound environment. When my mentor first introduced me to this topic, I found it incredibly interesting in the fact that it can be used in many ways and areas of research including muscle injuries, diabetic wounds, and cancer research. It also felt like a great fit for a project as I had always been interested in our lab’s research from a bioinformatics and computational perspective.
Throughout the project, my mentor encouraged me by giving me different directions to explore. This gave me a lot of freedom in what I could do with my project and made it a lot more fulfilling to work on. For most of my project, I worked mostly alone and in this I found that computationally driven projects are challenging in the fact that there is a lot of different methods and directions to go in and it is hard to determine where to start. To me this highlighted the importance of reading literature and methods as it helped me gather some sort of ground before starting my work. In the past, reading scientific papers had always been difficult and hard to read, but for my project it became interesting and gave me many new ideas and paths to explore.
After reading literature and analyzing code, I was able to produce spatial feature plots of macrophage biomarkers that gave us a better look at pro- and anti-inflammatory responses in mouse muscle injuries. Spatial feature plots overlay the expression of one gene over the tissue histology. Each spot had a specific color corresponding to the levels of gene expression in that location. To do this, I used bioinformatic packages built in R to construct these plots using a public spatial transcriptomic dataset of muscle injuries from mice. For pro-inflammatory biomarkers, the expression was localized to the wounded area while the anti-inflammatory biomarkers were spread throughout the tissue sample. Over time, an increase was observed in anti-inflammatory expression by 7 days post injury.
My future plan for this project is to perform the spatial transcriptomic data collection of muscle injuries in our lab rather than using a public data set. As this data set only spans 7 days post injury, much of the healing process is masked. Wounds in mice generally take around 20 days to heal so the current data set misses a large amount of wound regeneration. In future experiments, I would like to examine increased time points and days post-injury. I would also like to continue to improve the resolution and analysis of this dataset to improve the visualization of gene expression in the tissue.