top of page

Data with destiny

Imagine you have discovered a hit compound to treat drug addiction. Before you drop $1 billion on its development, wouldn’t you want to handicap the odds of its success as best you can? As we well know we do this poorly, but as the age of machine learning has taught us, new or overlooked sources of data can be transformative.

Some issues to consider: Is the target for the drug expressed in the NORMAL tissues that later are diseased? For example, if your drug is for depression, is the target expressed in the tissues of the brain responsible for mood? If your drug treats colon cancer, is its target expressed in the normal colon prior to the development of cancer?

These are more complex questions than they first seem. One view might be that diseases in those tissues arise BECAUSE a drug target that was poorly or not expressed aberrently increased its expression. In this case, as study of the normal tissues might show no or inverse correlation between the expression of a drug target in the prior healthy tissues and the success of the drug. On the other hand, one could argue that absence of the drug target from the prior healthy tissues portends poor to absent expression or function of the target in the diseased tissue, leaving nowhere for the drug to target that tissue. In this alternative case, absence of the drug target in the healthy tissue even when it is overexpressed in the diseased tissue might predict attrition down the line.

Kumar and associates carefully looked into this question a few years ago. This was also a surprisingly difficult task, since even the definitions of which cells comprise a specific tissue or which tissues are connected to a disease are debatable. Still, they persisted and obtained a preliminary answer: the expression of the target in the normal tissues might predict success down the line.

These are investigators after our own heart, as GeneCentrix makes tools precisely to combat attrition by revealing drug target expression in normal human tissues. Why not check the significance of the expression of your target in the expected disease tissues right at the start of your drug discovery effort?

Single Post: Blog_Single_Post_Widget
bottom of page