Bio leap InnoCos Europe, Paris

DrugDiscoveryDoneDifferentlyDavidL.Pompliano,PhD

CEO

What we do

BioLeap delivers custom-made, pre-clinical drug leads and candidates, in collaboration or as a service.

Enabled by computational fragment-based methods, we design novel compounds of unrestricted chemical diversity that bind to their target with predictable affinities.

By minimizing unproductive guesswork, we achieve pre-clinical milestones in shorter times and provide better value compared to current methods.

2

Many targets do not yield to the “process”

3

Targets Validated Targets

Hits Leads Candidates

>360 Predicted essential

~160 26

KO’s 70 HTS’s Medicinal

chem

5 0

Combi chem

The GSK Antibacterial Experience

Source: Payne, Gwynn, Holmes & Pompliano Nature Rev. Drug Discov., 2007 6, 29-40.

Lead identification and optimization is a time-consuming and inefficient process with low probability of success

There has got to be a better way.

BioLeap changes the approach from trial-correlate to design- confirm.

BioLeap’s hypothesis-driven design process

reduces compound attrition.

4

Fragment-based ligand design

1.  Find small, but highly specific, fragments 2.  Link them together (synergistic binding)

HT screening hit: asking too much all at once

Tight-binding drugs are composed of weak-binding fragments

6

TheBioLeapTechnology

ChemicaldiversityDesign‐centeredprocess

Predic@verankingofcompounds

7

Insufficient chemical diversity for screening (or an unwillingness to work on weak hits)

8

Chemical diversity: combinatorial chemistry with fragments known to bind to the target

Chemotype substitution sites for Imatinib # of chemical moieties

# of substitutions

per site

# of sites

= Possible Combinations

100 3 5 = 2.4T

100 3 4 = 8.1B

100 3 3 = 27M

BioLeap custom-builds novel-structure ligands from fragment building blocks that calculations show already bind to the target.

We DON’T screen !

Fragment Binding Map

BioLeap’s technology enables our chemists to expand their role as drug designers

Constrained Fragment Annealing

protein- centric

ligand- centric

BioLeap’s 3D Design Tools

Drug Designers

We create a map of where, and with what affinity, small chemical building blocks bind

11

Immerse protein Anneal µ

Isolated binding sites revealed

Lowest free energy, highest affinity site

A thermodynamically-principled model upon which to frame molecular design hypotheses

movie

Tools to rapidly assemble diverse fragments into novel compounds of predictable binding affinity

12 Chemists know which compounds to make next !

Predictable binding using BioLeap’s in silico annealing process

J.Med.Chem.2002,45,2994‐3008

No predictability using conventional docking

A Critical Assessment of Docking Programs and Scoring Functions Gregory L. Warren,*,† C. Webster Andrews,‡ Anna-Maria Capelli,# Brian Clarke,| Judith LaLonde,†,§ Millard H. Lambert,‡ Mika Lindvall,^,b Neysa Nevins,† Simon F. Semus,† Stefan Senger,^ Giovanna Tedesco,# Ian D. Wall,| James M. Woolven,^ Catherine E. Peishoff,† and Martha S. Head† GlaxoSmithKline Pharmaceuticals, 1250 South Collegeville Road, Collegeville, Pennsylvania 19426, GlaxoSmithKline, Five Moore Drive, Research Triangle Park, North Carolina 27709, GlaxoSmithKline, Centre via Alessandro, Fleming 4, 37135, Verona, Italy, GlaxoSmithKline, New Frontiers Science Park, Third Avenue, Harlow, Essex CM19 5AW, U.K., and GlaxoSmithKline, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, U.K. Received April 17, 2005 Docking is a computational technique that samples conformations of small molecules in protein binding sites; scoring functions are used to assess which of these conformations best complements the protein binding site. An evaluation of 10 docking programs and 37 scoring functions was conducted against eight proteins of seven protein types for three tasks: binding mode prediction, virtual screening for lead identification, and rank-ordering by affinity for lead optimization. All of the docking programs were able to generate ligand conformations similar to crystallographically determined protein/ligand complex structures for at least one of the targets. However, scoring functions were less successful at distinguishing the crystallographic conformation from the set of docked poses. Docking programs identified active compounds from a pharmaceutically relevant pool of decoy compounds; however, no single program performed well for all of the targets. For prediction of compound affinity, none of the docking programs or scoring functions made a useful prediction of ligand binding affinity.

GSK molecular modelers conclude that computational methods are not predictive

15 J. Med. Chem. 2006, 49, 5912-5931

In a blinded test with big pharma, BioLeap correctly ranked 87% of predicted binding affinities

16

-5

-4

-3

-2

-1

0 -50 -40 -30 -20 -10 0

Expe

rimen

talpIC

50

PredictedFreeEnergy

BioLeap controls key attrition factors

17

•  Biology: Target affinity and selectivity

•  Developability: Physicochemical properties

For MW, lower is better:

Source: J. Med Chem. 2003, 46, 1250-6.

BioLeap’s methodology is target-class independent

18

•  Kinases –  Mapkap-k2 (5 variants) –  P38 (3 variants) –  cAbl (2 variants) –  Ckit –  PhoQ Histidine kinase –  Proprietary kinases (3) –  JAK2/JAK3

•  Proteases and Hydrolytic Enzymes

–  Elastase: PPE, HNE serine proteases –  Peptide deformylase –  T4 lysozyme –  peptidyl t-RNA hydrolase

•  Nuclear Hormone Receptors –  ROR-alpha –  LXR

•  Oxygenases/Reductases –  Dihydrofolate reductase –  CpI hydrogenase –  Cox1/Cox2 –  IDO

•  Receptors –  EPO receptor –  NOGO

•  Macromolecular Interactions –  Protein/DNA complex –  P53/MDM2 –  BPTI (trypsin proteinase inhibitor) –  FABP4 –  Fcrn (peptide mimetic)

•  Other Classes –  NS5B RNA polymerase –  M2 proton pump –  Amino transferase –  Keap1 –  Arginase

In silico validation vs. known ligand Results confirmed experimentally, or in progress

BioLeap transforms economics of drug discovery

19

Comparison of Time and Resources Required to Produce a Development Candidate

BioLeap

•  150 compounds •  30 months •  Minimal infrastructure •  Library independent •  Broad diversity

Pharma

•  2,000 compounds •  48 months •  Big infrastructure •  Library dependent •  Limited diversity

Year 1 Year 2 Year 3 Year 4

3 Cycles Design / Test

2 Cycles Design / Test

Safety pharm.

HTS Assay Development Reagent Prep

HTS Hit Confirmation

Hit to Lead Chemistry

Lead Optimization Safety pharm.

Lead Candidate

Lead Candidate

Custom-built, ligand-efficient compounds with a past and a future.

•  Targets for which HTS methods have failed to produce new lead compounds

•  Targets where lead optimization efforts have stalled for a lack of understanding of the structure-activity relationship in the lead series

•  Expand/bust a patent

•  Develop a fast follower of an early stage clinical compounds

20

Extras

21

Leads are hard to find, and then the trouble starts

22

Discovery~8y Development~5.5y

Findingalead

LeadoptoproduceDC

“ValleyofDeath”

• Clairvoyance• Tenacity• Regulatorystability

What if the ideal position of the fragment is not consistent with chemical synthesis?

23

Only two obvious connections

Ligand-centric design: force constraints

24

Constrain link to amide

Constrain bond to N-atom and fuse

Design ligand Calculate FE with applied constraints

Constrained Fragment Annealing

Alternative designs are possible: choose based on ranking and synthetic feasibility

25

Control physicochemical properties and mode of target engagement

Accessvastchemicaldiversitybylinkingcombina@onsof

op@malfragments

Non‐obviousideasforleadop@miza@on

Circumven@ngtheSARparadox,avoidingpenal@esfrom@ghtly‐boundwaters

Automatedsearchforfragmentssa@sfyingbondgeometries Designforselec@vity,reduced

muta@onresistance

Exploitmoie@eswithstronginterac@onswiththebackboneorconservedaminoacidresidues

The goal of HTS

27

+