Genomics and AI: Cancer Research, Limits and Privacy
Why in the news
AI is speeding up genomics and cancer research, while doubts about accuracy, ethics and data safety grow.
About genomics
- Studies a genome’s structure, function and sequence, and how genes link to the phenotype.
AI advances
- Faster gene analysis enables broader studies.
- Johns Hopkins University (March 2024): machine learning examined 1,200+ ‘junk DNA’ elements, opening routes for therapy, diagnosis and screening.
- Gene Box startup uses AI for genetic predispositions and personalised advice.
Limits
- Genetics drives only about 30% of characteristics; environment and lifestyle dominate.
- Findings often labelled ‘of unknown significance’; Alzheimer’s risk genes do not guarantee the disease.
- Predictions on depression or heart disease need tougher guidelines.
Privacy concerns
- Nucleus (backed by Peter Thiel and Alexis Ohanian) wants genomes on every iPhone.
- October 2023 breach: 6.9 million users affected, $30 million fines, board and staff largely lost.
- Withdrawing data is hard; no HIPAA cover for consumer genetic data.
Exam angle
- Terms: genome, phenotype, junk DNA, HIPAA.