BIO/ML Researcher: "AI Cannot Cure Cancer" - Favour Igewzeke

Описание: In this episode, I sit down with Favour Igwezeke, a computational genomics researcher who built machine learning models to predict tumour aggressiveness using DNA sequencing data.

Favour led an end-to-end genomic analysis of meningioma tumours using whole-exome sequencing data from TCGA and cBioPortal, developing computational pipelines to measure tumour mutational burden and genomic instability. His work revealed significant genomic differences between tumour grades and achieved an AUC of 0.879 in predicting tumour severity using machine learning models like Random Forest and XGBoost.

We discuss how AI is transforming cancer research, how tumour DNA can reveal hidden biomarkers, and what the future of computational medicine looks like.

This conversation covers:

• How AI can predict tumour aggressiveness
• Using genomic data to discover cancer biomarkers
• Building machine learning models for real clinical problems
• Tumour mutational burden and genomic instability explained simply
• Lessons from working with TCGA and real patient sequencing data
• The future of AI in healthcare and cancer diagnosis
• Advice for students and researchers entering computational biology

This episode is for anyone interested in AI, machine learning, genomics, healthcare innovation, or the future of cancer diagnostics.

EP 1

Guest: Favour Igwezeke
Field: Computational Genomics | Machine Learning | Cancer Research

Favour's linkedin:   / favourokechukwu  

Paul-Simon ( The Host ):
Paul-Simon's Website: https://www.paulsimon.engineer/
Paul-Simon's X: https://x.com/ptbthefirst
Paul-Simon's Linkedin:   / paul-simon-emechebe-babbb726a