In contrast to target-based strategies, phenotypic drug discovery does not rely on knowledge of a specific drug target or a hypothesis about its role in disease, but instead focuses on finding a gene or small molecule that corrects a cellular characteristic that is specific to disease.
Phenotypic screening has recently led to discoveries of first-in-class drugs with novel mechanisms of action. More importantly, these kinds of drugs discovered via phenotypic screens seem to be approved at a higher rate than traditional target-based screening. It's the anti-attrition screen! Due to this success, interest in this strategy experienced a renaissance over the past 5 years, but the pendulum is already swinging back as the central challenge of phenotypic screens–target deconvolution–remains as thorny as ever.
In addition, the new technologies that helped the surge in the first place created a paralyzing data deluge. Major advances in high throughput technologies have resulted in immense phenotypic datasets. However, these recent advances in automated screening technology have resulted in a dilemma: Experimental data can be generated at a much faster rate than researchers can possibly analyze and integrate them. This inefficiency makes it difficult to conduct certain types of experiments on a large scale, wastes investments, and could potentially delay the drug discovery process.
Well, just as computation saved cryoelectron microscopy from the fringes of structural biology, scientists are gathering in New York City this year to see if it can do the same for phenotypic screening. This symposium will provide a snapshot of the current state of computational methods used in phenotypic screening and novel in silico approaches, and include discussions of deep learning, AI, functional genomics, chemical screening, systems biology, target deconvolution, biomarkers, and toxicity. GeneCentrix will be there to show how our platform, Historeceptomics, both results in massive reduction in the dimensionality of the big phenotypic screen datasets AND enables target deconvolution. Don’t miss it!!