fragments and beyond

Off-the-Shelf Fragment Library

Fragment program at ChemDiv is aimed to address another approach in drug discovery – “fragment-based” drug design (FBDD). The immediately available fragment library includes 10,000+ structures. Those were selected from 1.2+ Mio collection and produced as the result of internal design & synthetic effort in consideration of the following guidelines:

•    Widely accepted Astex Rule of Three[1] (MW ≤ 300, H-bond donors ≤ 3, H-bond acceptors ≤ 3, cLogP ≤ 3)
•    The tagged fragments may contain only C, H, N, O, S, P, F, Cl, and Br atoms
•    The tagged fragments must be chemically feasible for follow up linking and/or expansion
•    The tagged fragments comply with internal medicinal chemistry filters (this way undesirable moieties, such as Michael acceptors, can be removed upfront)
•    Add-ons of structures with fragment-“friendly” features, such as:

-    NMR screening compatibility (chemical shifts of protons to be in the 0 – 12 range evenly; possession of both alkyl & aromatic protons, or “marker atoms”, e.g. F)
-    X-ray screening compatibility (heavy atoms)

As with any other “classical” small molecules library, the data base with immediately available fragments can be provided (as MDL structure-data files (SDfiles) or ISIS Database (db) files) for follow up selection and ordering. The fragments can be dispensed in mg or micromolar amounts and delivered in a variety of formats (plates or vials, as dry powders or DMSO solutions).

Custom Fragment Generation

At ChemDiv, we propose rational selection and/or design of drug-like templates with follow-up synthesis small series (5-10 molecules) of lead-like fragments around those. The main principle of such template design / selection consists in the use of knowledge database of known ligands (kinase, GPCR, ion channels, etc.) and enzyme effectors as the prototypes. The novel designed templates undergo through RECAP (Retrosynthetic Combinatorial Analysis Procedure) method of dissection[2] – fragmenting molecules around bonds which are formed by common chemical reactions, – thus generating novel synthetically realizable fragment-like space. The standard set of fragment selection guidelines is deployed to shape up the final proposed fragment space.

The advantage of this approach is that initial templates are “fragmented” at several pre-defined bond types all of which are amenable to combinatorial chemistry. Therefore, the created novel fragments represent direct precursors of building blocks for combinatorial library synthesis. To avoid unnecessary simplification of some privileged templates (e.g. dissection of biphenyl fragments), in some cases we use the modified RECAP approach – it considers not only chemistry-derived set of rules, but also distinctive structural features of such privileged templates. For example, we propose several bond types to be left intact – all mono- and biheterocyclic structures, benzylheterocycles, spirocyclic fragments, biphenyl and diarylmethane fragments (and their heterocyclic bioisosteres), also certain ring fragments.

As a result of our internal novel design process, we constantly generate and prosecute on hundreds of novel templates each year. In parallel with this process we also constantly add novel tractable (synthetically realizable) fragments into our pool of such synthetically realizable chemistry. Upon your request this up-to-date data base of such virtual fragments can be provided for further selection and synthesis on demand. We can also provide the data base of validated scaffolds and generate tractable fragment space for further production for those that will be selected.

Expansion on Active Fragments

In this research segment we offer tools and services to address the need of a) fragments assembly (linking) into “classical” small molecules, and b) custom expansion on active fragments.

Within our collection of building blocks we made available a sub-set of linkers – specific building blocks that have two (or in some cases more) chemically orthogonal functionalities on a inert, flexible or rigid core. Such functional (reactive) groups can be used to connect (or “link”) active fragments to each other thus forming a new small molecule. Our data base of immediately available “linkers” includes thousands of structures which can be delivered in various amounts. By tagging such bi-functional building blocks as linkers we were guided by following considerations:

•    Various distance between reactive groups in a building block core;
•    Various 3D-spacial position of reactive groups in a building block core;
•    Various conformational flexibility of inert part in a building block core;

Additionally, the offered bi-functional building blocks (“linkers”) can also be deployed to transform active fragments into mono-functional building blocks.

We also provide services on custom expansion on active fragments. Specifically, we are in the position to analyze the fragment-target interaction mode and select / design appropriate linkers for follow up expansion on selected fragments while achieving desirable special conformation of the final molecule. At this stage we can also:

•    introduce bulky substituents;
•    rigidify structural elements (e.g. fixing rotatable bonds in space, etc.);
•    introduce stable isotopes or heavy atoms into fragment core;

As such, the assembled molecules will comprise various a) dipole magnitude, b) distance between defined marker atoms, c) spacial orientation of original fragment(-s) in a new molecule.


[1] Carr, R.A.E. et al., Drug Discovery Today, 2005, 10, 987-992.

[2] X. Q. Lewell, D. B. Judd, S. P. Watson, M. M. Hann, J. Chem. Inf. Comp. Sci. 1998, 38, 511-522.

Explore other libraries:

Discovery
Targeted Diversity
Focused Libraries
Peptidomimetic Libraries