Metagenomics is the study of microbial genetic material taken directly from an environment. One of the many achievements of metagenomic study is the growing understanding of the microbes inhabiting the human intestinal tract and their correlations with various diseases, ranging from obesity to stomach cancer. The bacteriophage, crAssphage, has been recently discovered within human fecal samples. It is present in half of the human population sampled for it and is predicted to be six times more abundant than all other annotated phages combined. CrAssphage is predicted to infect bacteria of the genus Bacteroides, which has been implicated in various gut related human diseases such as diabetes. Even though the function and host of this virus remains unknown, identifying which samples contain crAssphage will aid in the characterization of this phage by identification of the ubiquity, locations and associations with health and disease. This task is difficult due to phages’ rapid mutation rate, which prevents assembled reads from matching the reference sequence, and the low likelihood that the entire genome of crAssphage will be present in the sample. A method of overcoming these obstacles, would be to calculate the frequency that crAssphage’s k-mers appear in a fecal metagenome. A random forest classifier was utilized to determine which of crAssphage's k-mers are ideal predictors for its presence. The goal of this project is rapidly identify whether or not a fecal metagenome is positive for the presence of crAssphage by determining the abundance of its predicting k-mers.
