ABSTRACT

Clustering and Detecting Nucleotide Modifications on 18S Yeast (S. cerevisiae) rRNA

 

Shreya Mantripragada (Monta Vista High School) and Alejandra Duran (Colegio Santa Francisca Romana)

 

Advisor: Andrew Bailey (University of California, Santa Cruz)

DNA and RNA are essential components of all cells and are crucial for survival and growth. Nucleotide modifications are vital for well-regulated DNA and RNA function. During our research, we explored nucleotide modifications on Yeast (S. cerevisiae) rRNA (18S subunit), which are necessary for ribosomal diversity and translation. Using a high-dimensional dataset containing the nucleotide modifications in 18S rRNA, analyzed through nanopore sequencing, we hypothesized that we could use clustering algorithms to find patterns in nucleotide modification occurrences at a single-strand resolution level. First, we researched numerous clustering algorithms and found the optimal values for their parameters. We then implemented these algorithms on our dataset and visualized the produced clusters after reducing our high-dimensional dataset with t-SNE Dimensionality Reduction. KMeans and Spectral Clustering algorithms produced the clearest clusters on our data and looking at each produced cluster, we were able to conclude the probability of a certain modification combination occurring on a given Yeast 18S rRNA strand. Since our work is one of the first to analyze rRNA at a single-strand resolution level, our results depict the function of eukaryotic rRNA and its fixed modification patterns that differ from strand to strand. Our project presents an initial approach in identifying modification patterns in a given rRNA strand, patterns that could be useful in understanding functions of the rRNA strand located in the human genome.