Selcia Fragment Library Aimed at broad portfolio of targets Currently ~1,3XX compounds (30 mM stocks in DMSO) Selection & design by ‘Rule of 3’ criteria1) >20% fragments are substructures in marketed drugs Good quality control Removal of reactive compounds (Medicinal Chemistry perspective) Measured solubility criteria (measured aqueous solubility >1mM) Purity criteria (>95%) by 1H-NMR High degree of diversity Majority are singletons (Tanimoto coefficient 0.9) Diversity coefficient 0.752 (highest value compared to 15 other commercial fragment libraries)3 Low overlap with commercial fragment libraries 1Congreve et al. Drug Discovery Today; 8 (2003) 876-877; 2Turner et al. J Chem. Inf. Comput. Sci. 1997, 37 pp18-22; 3 http://www.cambridgemedchemconsulting.com/resources/hit_identification/fragment_collection_profiles.html

The library was initially assembled using compounds from Selcia’s compound collection and commercial compounds obtained from fragmentation of known biologically active compounds. The compounds were selected based on ‘rule-of-3’ criteria1, chemical attractiveness and potential for expandability (i.e. availability of near neighbours). Fragments that met the ‘rule-of-3’4 and calculated solubility criteria (>1mM) were prioritised for inclusion. The diversity of the collection was enhanced by the inclusion of both commercial and non-commercial, custom synthesised fragments.

Selcia Fragment Library Characteristics The characteristics of the 13xx compounds (those that met the QC criteria) were calculated using ChemAxon tools Median = 160 Median = 45 Median = 1.2 Median = 1 Median = 12 Median = 2 Median = 2

Library Quality Control Due to the high solution concentrations of compounds used during a fragment screen, it is important to ensure those compounds are soluble and of good quality. Collection quality assessment and solubility were measured by performing detailed LC-MS and 1H-NMR; those that did not meet the set criteria were excluded. The compounds were dissolved as 30mM stocks in DMSO. Calculated Solubility: The aqueous solubility of the parent compound (i.e. non-salt) was calculated using WSKOWWIN (part of EPI SuiteTM v4.0 available from http://www.epa.gov/opptintr/exposure/pubs/episuitedl.htm). A calculated solubility criteria of >1mM was used. Measured Solubility: Solubility was measured using a turbidimetric method. The fragments were diluted to 1mM in 35mM Hepes pH7.8 (3.3% final DMSO in clear 96-well plates. The plates were sealed and shaken for 3h at 25oC. The absorbance at 650nm was measured using a SpectraMaxM5 (Molecular Devices). If the background subtracted A650nm was ≥0.01 in both duplicates the compounds were scored as insoluble, those <0.01 were scored as soluble. Reactive Compounds: All fragments were reviewed by Selcia’s Medicinal Chemists and any ‘reactive’ compounds were removed (e.g. electrophiles). LC-MS: LC-MS analysis of the library was performed on a Quattro LC (Micromass). Purity was assessed by UV. NMR Analysis: The library was analysed by 1H-NMR using a Bruker Avance 500MHz system. Spectra for each compound were visually assessed for compound integrity and % purity by two independent Medicinal Chemists. Frequent Hitters: Frequent hitter/problem compounds were identified through use of the collection and removed. Long Term Solubility: Compounds failing longer stability QC (1 year in DMSO) were removed.

the majority of the fragments in the collection are singletons, indicating a diverse collection of fragments, and there is very little overlap between the Selcia Fragment Library and other commercially available fragment libraries.