Iterative Focused Screening
An alternative approach to high-throughput screening (HTS) is to apply iterative focused screening (IFS) for instances where targets do not permit an HTS campaign, a method practiced by Merck Frosst Canada (www.merckfrosst.com).
The rationale is to use a small subset of postulated actives, enriched with inactives and Random Forest, a statistical modeling tool, to select a biased test set of molecules from a large compound library. This results in higher quality data by testing samples at three doses followed by confirmation of actives before the next round of IFS.
Were one of the smallest Merck basic research sites in the world but we are still 200 scientists strong, notes Christine Brideau, senior research fellow, biochemistry and molecular biology. We have our own internal drug discovery programs here. We develop HTS assays and perform most of the screening except for ultrahigh-throughput screening, as this is centralized in the U.S. At IBC we will explain how our methodology came about. When you are doing high-throughput screening, you are randomly screening compounds. The more you do, the more likely you are to find a hit. But, how many do you need to test to get results?
Our method rapidly identifies true hits and is more successful than others to reduce the number of screening iterations and test samples. Results from two examples of IFS will be shown at IBC and compared with real HTS data to demonstrate how our approach increases the hit rate and is successful in identifying novel lead structures.
Screening technology has moved forward so much that we can screen millions of compounds in a short time. The only way you can do that is to miniaturize and automate the assays. But, not every assay is compatible to miniaturization or automation. Another issue is that certain targets are unattainable in that not enough information is available about the biology of some targets, limiting the tools available to develop HTS assays.
Using IFS allows you to kickstart the process of filtering through, first finding the molecules that do interact with the target, explains Brideau. And it allows you to eliminate from the beginning the molecules that have a lower probability of interacting with the target. The first time wont necessarily give you the best leads, but youll get a smaller number of actives to initiate the next round of IFS, until new leads are found. And you can adjust the algorithm accordingly. Using IFS we can test fewer compounds but in replicates, where in HTS you can only afford to test 1 million compounds in singlet. IFS is more robust than HTS.
Over one-half of the protein-encoding genes in the human genome have no known function. These molecules represent an unprecedented opportunity for novel target identification. In conjuction with genetic analysis in mammalian cells, the Genomics Institute of the Novartis Research Foundation (GNF; www.gnf.org) has employed high-content imaging to interrogate complex molecular phenotypes, such as cell migration, nuclear translocation, and cell cycle at the level of the genome.
GNF is undertaking the first industrial and comprehensive attempt to unveil the function of each gene in the genome, using both cDNA and siRNA to assess the affect of overexpression and gene knockout, respectively, in phenotypic cellular assays.
Were actually doing genetics in mammalian cells, says Sumit K.Chanda group leader, genomics. Were looking at one gene at a time, and our challenge is modulating one particular cell phenotype at a time.
To this end, GNF, with the aid of Novartis, has compiled and arrayed in excess of 20,000, unique, full-length human cDNAs in mammalian expression vectors in 384-well plates, each well containing a unique gene. Custom-built plate and liquid-handling robotics screen for effects in a variety of complex cellular assays. Current capabilities allow screening of more than 150,000 genes per day.
A lot of the latest technology is siRNAs, how we set up and how we analyze data becomes an automated process, Chanda notes. The advantages are tremendous and obvious. It allows us to systematically knock out every gene in the genome. At the end of the assays, what you get is a list of genes, you take that list and triage it. What you do with this information will make or break the target identification project at this point. It is an incredibly powerful methodology, but as a tool, it is only as good as the person using it. And it is a tool, not the final answer.
Readouts include luminescence, fluorescence, and recently through collaborative efforts with local biotech and academia, GNF is performing high-content imaging screens for distinct cellular events including differentiation, cell cycle arrest, apoptosis, protein translocation, membrane ruffling, neurite outgrowth, and others. These efforts have enabled the identification of many novel effectors in even the most well-studied signaling pathways, as well as ascribing novel function to old genes.
One thing plaguing the field right now is off-target activity, notes Chanda. This is the greatest source of false positives, and Ill discuss strategies to minimize false positives.