In August this summer, Professor Mathias Uhlén and colleagues from The Human Protein Atlas at KTH, took a step closer towards personalized medicine in oncology. By analysing protein expression in 17 different types of cancer, using big data mining of samples from 8000 patients, the team revealed expression patterns that can be linked to patient survival and hence be of importance in personalized therapy as well as in developing next generation protein pharmaceuticals to fight cancer.
Obviously, the quest is not new – for decades, researchers have been wrestling the question of what it is that trigger a cell to become tumorigenic and go berserk with number of cell divisions. What is unique with this study however, is the broad approach which is possible in an era when research findings are publicly available on the web. Many of these original findings are from the Human Protein Atlas, others from other part of the research community.The Human Pathology Atlas is publicly available and can be used by researchers as well as companies from the pharmaceutical industry. The atlas contains over 900 000 graphs comparing patient survival with the expression levels of different genes.
Perhaps not surprisingly, results reveal that genes involved in DNA replication, cell division and apoptosis – programmed cell death, seem to be expressed at higher levels in cancer cells than in normal cells. A set of genes can also be linked to poor survival, those genes are mainly responsible for cell division and growth.
Over 2000 genes was shown to have an effect on patient survival depending on which cancer type the patient had and where in the body the tumour was located. Another set of some 2000 genes were also identified to be potential targets of tumour growth – however, affecting most of those targets would be associated with severe side effects for the patient and therefore are not suitable as drug targets. The funnel of potential drug targets narrows down however, as the team found 32 genes that were expressed in over 80 percent of the tumours regardless of cancer type. It is not a wild guess that those proteins will be meticulously examined in the future, in search of improved cancer treatments, as a step closer towards personalized medicine in practice.