Navigating the DNA encoded libraries chemical space

Abstract

DNA-encoded library (DEL) technology is a novel ligand identification strategy that allows the synthesis and screening of unprecedented chemical diversity more efficiently than conventional methods. However, no reports have been published to systematically study how to increase the diversity and improve the molecular property space that can be covered with DEL. This report describes the development and application of eDESIGNER, an algorithm that comprehensively generates all possible library designs, enumerates and profiles samples from each library and evaluates them to select the libraries to be synthesized. This tool utilizes suitable on-DNA chemistries and available building blocks to design and identify libraries with a pre-defined molecular weight distribution and maximal diversity compared with compound collections from other sources.

Fig. 1. eDESIGNER concepts.

a Description of the overall eDESIGNER process: reactions and building block types (BBTs) are combined to generate all possible eDESIGNs, then eDESIGNs are merged into libDESIGNS when they can be implemented together in the same experimental conditions. Only libDESIGNs meeting specific criteria survive and their diversity is analyzed through a sample of 10,000 compounds per libDESIGN. b eDESIGN and libDESIGN objects: BBTs and reactions are combined in graph objects in which edges represent reactions and nodes are BBTs. libDESIGNS combine several eDESIGNS with the same topology and compatible reactions. c Functional group (FG) and BBT objects: a functional FG is a handle that can be used to link BBTs through reactions. Some FGs are exposed for reactions, and some are protected and need deprotection reactions to become exposed. The null FG is added to simplify the code. A combination of FGs defines a BBT. d Reaction object: connecting reactions connect two FGs to link two BBTs eliminating the FGs used (entry 1) or creating a new one (entry 2). Deprotection reactions transform an FG into a different one without addition of mass (entry 3) or incorporating a scaffold (entry 4). e Iterative process to generate an eDESIGN: a growing eDESIGN, characterized by its list of FGs, can then further incorporate new BBTs using additional reactions, convert protected FGs in exposed FGs, or add a non-coded scaffold.

Fig. 2. eDESIGNER design examples.

a 2-cycle libDESIGN. b 3-cycle libDESIGN.

Fig. 3. Building block analysis.

a Effective heavy atom distribution of BBs by source (magenta: external BBs, cyan: internal BBs). b Number of building blocks with respect to the multiplicity in their functionality (cyan: monofunctional, magenta: bifunctional, orange: trifunctional). c Analysis of BBs used by eDESIGNER in the two-cycle library designs by BB multiplicity. d Analysis of BBs used by eDESIGNER in the three-cycle library designs by BB multiplicity.

Fig. 4. Molecular property profiles.

Property distributions for calculated properties. Sample of 10,000 compounds enumerated from each two-cycle libDESIGN (magenta), three-cycle libDESIGN (cyan), previously synthesized libraries (ADEL collection, orange), and 140,000 drug-like compounds belonging to the Lilly diversity cassette (LDC collection, yellow).

Fig. 5. Diversity analysis.

a Average distance of two-cycle libDESIGN library samples to ADEL. b Average distance of two-cycle libDESIGN library samples to LDC collection. Scores are represented versus library size and lines connect enumeration sets (10,000 compounds each) drawn from samples prepared using internal BBs only to samples using all types of BBs. c Spread design using the ADEL and LDC collections as reference for two-cycle libDESIGNs using internal BBs only. Scores are represented versus library size.