Author(s): George Benson
The ability to provide initialed or individualized medicine to cancer patient largely depend on the ability to discover the different cancer subtypes. As advanced biological knowledge databases and multi-level omics datasets continue to accumulate, there is urgent need to integrate the current extensive biological knowledge with the omics data to decipher a natural mechanism that is responsible for deadly diseases. The researchers sought to recommend an integrative sparse K-Means (IS-Kmeans) method to facilitate discovery of subtypes of diseases and cancer drugs mostly guided by past biological knowledge. Additionally, to achieve quick optimization, they relied on an algorithm that used an alternating direction method of multiplier (ADMM).