Fragment-based lead discovery (FBLD) has undoubtedly become an invaluable resource for medicinal chemists in the identification of lead candidates.
Since an efficient fragment evolution requires structural data of protein-ligand interactions, a variety of biophysical methods are applied in a cascade approach as a set of prefilter prior to structure determination. Systematic studies suggest that the cascade approach significantly reduces the hit rate compared to a direct crystallographic screening.
The application of X-ray crystallography as a primary fragment screening method has been underutilized due to the limited availability of a robust system for soaking experiments and solubility issues of fragments. One of major bottlenecks in the experimental setup for crystal soaking is the sensitivity of protein crystals. We overcome this limitation using the SmartSoak® technology, which stabilizes protein crystals leading to high-performance soaking systems. No trial-and-error optimization of the soaking system needed.
In case studies, we demonstrate the advantages of the application of crystallographic fragment screening as a primary screening strategy for the hit identification. The resulting structural data can now be utilized to explore the chemical space on scale using computational methods.
While the amount of publically available protein structures substantially grows, the software ecosystem for life scientists interested in gaining knowledge from structures is still mostly based on commercial, monolithic modeling software. With increasing capabilities of web browsers with respect to visualization and interactivity, it is the right time to established web-based modeling systems. The aim is to make these systems reliable and easy to use such that students and ocasional users from life science have a low entrance barrier. About five years ago, we started the development of the ProteinsPlus web server. Based on the PDB visualization (NGL viewer) and query REST API, ProteinsPlus offers a series of structure processing services allowing to get quick insights from structural data. Today, ProteinsPlus offers 11 tools, the most recent addition is GeoMine, the first geometric search engine on the entire set of protein-ligand interfaces from the PDB. GeoMine allows to browse protein-ligand complexes by geometric features with typical response times in the range of tenths of seconds to minutes.
Finding molecules with completely new structures is of ever-growing importance. Lately, searching in vast virtual compound spaces is experiencing an unseen wave of success in this domain. The reason is simple: compound vendors have started to encode their reaction knowledge and offer billions of compounds that they can theoretically make as "compounds-on-demand". Based on robust reactions and highly suitable building blocks, quick and inexpensive synthesis with high success rates is possible. However, with just one 3-component reaction and 1000 reagents in each position, 1 billion compounds become accessible. Therefore, the traditional search for the needle in the haystack has turned into a search for many needles albeit in a mighty haystack.
In this webinar we explain what a chemical space actually is, how chemical spaces are built based on reactions and reagents. We show you how to explore such spaces by similarity searching, including a successful application example. You will see how traditional library searches are doomed to fail due to the combinatorial explosion, whereas our space concept has proven successful with spaces the size 1020 virtual products. Think about this number as an enormous haystack, but it's also loaded with many valuable needles. Last but not least, we will show you how we explore chemical spaces in 3D, with a novel docking strategy. This method has been applied successfully on a number of internal projects. Here we will first disclose some recent results.
The discovery that proteins and/or protein families of interest can be labelled selectively with chemical reagents resulted in an active research field that aims to profile the amount and activity of (a subset of) proteins in a relevant setting. For this purpose, a diverse set of so-called activity- and affinity-based probes (ABPs) have been developed nowadays. The molecular structure of these probes generally consists of a recognition element, which generates selectivity for the protein of interest, a reactive group or warhead, which facilitates covalent binding of the probe to the protein, and a reporter group, which is used for read-out purposes. The activity-based protein profiling field has matured over the past decades and these research tools are nowadays employed in both fundamental and applied settings.
During this webinar, I will introduce the basics of activity- and affinity-based protein profiling. I will describe the general design approaches for ABPs, as well as their applications. In the second part of the webinar, I will discuss novel combinatorial approaches to prepare probes and I will demonstrate how chemical probes can be combined with nanoLC-MS/MS and docking studies to map the binding sites of small molecules.