Dominik Glodzik's research at Harvard DBMI is focused on applications of statistical algorithms to understand, treat, and detect cancer early.
During his postdoctoral fellowship at the Sanger Institute in the groups of Sir Prof Mike Stratton and Prof Serena Nik-Zainal, he became an expert on detecting patterns of mutations in cancer genomes. Specifically, he pioneered using supervised machine learning methods to understand mutational processes in cancer. His most widely used algorithm is HRDetect (Nature Medicine, Davies and Glodzik et al., 2017). This algorithm identifies cancer patients with homologous recombination deficiency (HRD) from genome sequencing data, widening the population of patients eligible for therapies.
On-going work is focused on the following areas:
- machine learning methods for the prediction of therapeutic vulnerabilities from mutational patterns (mutational signatures) in cancer, using the principle of synthetic lethality.
- utilizing known mutational patterns to create more sensitive liquid biopsies for early cancer detection.
Cancer Res
View full abstract on Pubmed
Nat Cancer
View full abstract on Pubmed
Nat Genet
View full abstract on Pubmed
Blood Cancer Discov
View full abstract on Pubmed
Nat Commun
View full abstract on Pubmed
Nat Cancer
View full abstract on Pubmed
Cancer
View full abstract on Pubmed
Nat Cancer
View full abstract on Pubmed
Nat Med
View full abstract on Pubmed
Nat Commun
View full abstract on Pubmed