Chemical diversity the Good and the Bad

1
Chemical diversity the Good and the Bad

• We need to find the chemical space that overlaps
  with the space of TSC biology in order to find a
  drug for TSC.

Chemical Space
Biological Space
Overlap potential drugs!
2
Chemical diversity the Good and the Bad

• The larger the chemical space (i.e. the more
  diverse your library) the larger the likelihood
  of success (overlap of biological space and
  chemical space)
• Since we know little about the biological space
  of TSC the above statement is even more
  significant.
• This results in a huge Haystack in which we try
  to find a needle.

Unknown biological space
3
VLS Virtual Library Screening

• The Task Reducing the size of the Haystack

• We currently only consider commercially
  available compounds since our targets have not
  been validated.
• Currently 2.2 Million compounds are
  commercially available.
• At a cost of 4-12 US / compound, screening all
  compounds is cost prohibitive.
• Since we dont know how to revert the effect of
  TSC1 and/or TSC2 knockout, we need to explore the
  effect of small molecule binding to known members
  of the pathway (i.e. validating targets
  implicated in pathway which have x-ray crystal
  structure). We use VLS to find the best binders
  before buying these compounds for screening.

4
Reducing the Haystack

• We use various filters to take out the
  compounds that have no chance of overlapping with
  our biological space (especially since we test
  in vivo)
• - drug like molecules (Lipinskis Rule of 5)
• The compounds that make the second round are
  screened using Docking algorithms to reduce the
  numbers even further. This constitutes virtual
  library screening, since the theoretical results
  should be reflected by the activity of the
  compounds in experimental results.

5
How does VLS work?

• The Lock and Key mechanism of Small molecule
  interactions

Small molecule
Protein (Target)
Ideal Fit highest activity

• In VLS we rank the small molecules according to
  their fitness to bind our Target.

1. 2. 3. 4. 5. 6. 7. 8.
6
What is required?

• To screen a library of several hundred of
  thousands of compound requires enormous
  computing power.
• The Rothberg Institute has a user community
  (8000 members) that donates CPU time by running
  a screen saver which does the docking
  calculations when the computer is idle.
• This gives us computing power several times
  greater than the biggest clusters in the world.

7
D2OL Drug Development and Optimization
Laboratory

• This is a client-server based system which runs
  on all common platforms Windows, MacOS X, Linux,
  Solaris
• The client is downloadable from our Website
• www.childhooddiseases.org

8
The Screen Saver
9
Join a Team !
10
Statistics
11
Science

• So far the success has been limited

• We strive to improve and perfect the docking
  process to obtain 1000 fold enrichment of our
  libraries.

12
The Future

• Upon validation of targets DE NOVO drug design
• Either Ab Initio (based on 3-D structure of
  
• active site or allosteric binding site)
• Or Based on Lead Compound obtained from
  first round of screening. (Traditional
  Medicinal Chemistry)