Microbe-Cancer Pairs
This project on applying AI to rank and prioritise research on novel microbial oncogenesis is a joint initiative of Wits IDORI and Wits MIND.
The discovery of human papillomavirus (HPV) as the main cause of cervical cancer (CC) transformed public health oncology, leading to CC becoming the first cancer that could be potentially eliminated through infection prevention and treatment.
Today, microbes are estimated to contribute to 1 in 5 cancers worldwide, prompting growing research efforts to identify additional microbe-cancer pairs (MCPs) with similar potential for prevention and public health impact. The result is a vast and rapidly expanding body of biomedical literature containing evidence on numerous MCPs. Identifying which are most plausible, and which would have the greatest impact if targeted, requires careful analysis of the literature.
However, it is unlikely for any team to comprehensively review this volume of work, so we are developing a transparent and validated AI-driven pipeline that automates this process, handling literature search and retrieval, dataset refinement**, and the extraction and synthesis of key findings. By systematically evaluating the strength of evidence for each MCP and combining this with measures of cancer incidence and disease burden, the pipeline will generate interpretable plausibility and impact scores, enabling MCPs to be ranked in a way that can guide future research priorities and resource allocation.
Beyond MCP research, this pipeline is designed to be adaptable to a wide range of causal questions. Our goal is to make large-scale literature assessment more accessible, transparent, and reproducible - empowering researchers to engage with the full scope of existing evidence and thereby accelerating scientific progress.
A preprint of our dataset refinement component is available here.

More information
- Find out more about Wits MIND here: /mind/
- Contact the MCP team at: bruce.a.bassett@gmail.com
