CRCPred

CRCpred

CRCpred is a tool designed to predict the healthy or CRC status of an individual based on gut bacterial abundance profile.
Significance

Colorectal cancer (CRC) is a leading cause of death globally. Numerous studies have reported that the gut microbiome plays a significant role in the development and progression of CRC. The alteration in the gut microbiome associated with CRC presents an opportunity for early detection through microbiome-based diagnostics. Here, we developed CRCpred, which integrates machine learning and artificial neural networks to predict CRC using the relative abundance of gut bacterial species. CRCpred is built on a dataset of 1,728 metagenomic samples from 11 studies across eight countries, ensuring that the tool is robust and generalizable. It employs three classification methods, including XGBoost, Random Forest, and ANN to analyze and classify samples as either healthy or CRC. XGBoost demonstrated high performance, with an average AUC of 0.91 on the training and the validation sets. Random Forest achieved a mean AUC of 0.91 on the training and the validation sets, while ANN achieved comparable performance with an AUC of 0.89 on training and 0.84 on validation sets. CRCpred shows promise as an effective tool for microbiome-based CRC diagnosis.

The flowchart of the approach used in CRCpred is shown below.

MetaBioSys lab, Indian Institute of Science Education and Research (IISER) Bhopal, Bhopal bypass road, Bhauri, Madhya Pradesh, India (462066).

METABIOSYS