The human microbiome consists of complex interactions between microbiota, which play a significant role in human health. Sequencing data from human fecal samples offer a snapshot of the microbiome and have the power to suggest onsets of a variety of diseases. With many medical diagnostic tests being expensive, invasive, and limited to the disease of interest, an alternative method that is cheap and effective across many conditions can make diagnosis more accessible and allow individuals treat diseases earlier. Microbiome data has this potential, however as the interactions between microbiota are not deeply classified and complex, a machine learning model is an effective method to interpret this data and output medically relevant data. I trained a Random Forest Classifier on microbiome data from patients with and without colorectal cancer.
Diagnosing Disease with Human Microbiome Data
With the use of machine learning models, sequenced fecal samples alone can be used for preliminary diagnosis for a variety of diseases from colorectal cancer to diabetes.
Updates
Leave feedback in the comments!
Log in or sign up for Devpost to join the conversation.