Distributed Drug Discovery

1
Distributed Drug Discovery

• Indiana University Purdue University, Indianapolis

2

• Introduction
• PURPOSE
• OUTLINE
• DESIGN
• IMPLEMENTATION
• FUTURE WORK

3
Purpose

• Need for innovative and inexpensive drugs to
  treat diseases in the developing world.
• Developed world has the equipment and procedures
  in the pharmaceutical industry through the
  economic incentives.
• Developing countries do not have the required
  economic incentives for the pharmaceutical
  industry.

4
Concept

• Concept of Distributed Drug Discovery is that
  if simple, inexpensive equipment and procedures
  are developed for each of the core scientific
  drug discovery disciplines (computational
  chemistry, synthetic chemistry and biochemical
  screening), large research problems can be broken
  down into manageable smaller units and carried
  out at multiple academic sites throughout the
  developing and developed world.
• The coordinated and recombined results of these
  resources would inexpensively accelerate the
  identification of leads in the early stages of
  the drug discovery process.

5
Procedure

• A list of potential drug lead candidates could be
  split up into small batches for synthesis in
  academic laboratories throughout the world.
• Then the pooled molecule products can be tested
  by the distributed screening effort.
• The summation of these efforts, networked across
  the three core disciplines, becomes a powerful
  globally distributed drug lead process and
  solution.

6
Procedure (contd.)

• The synthetic laboratory experimental work is
  done in three main locations around the world
  including IUPUI.
• Enumeration software is used to create large sets
  of potential product molecules called virtual
  libraries.
• These in turn are rehearsed by students using a
  distributed approach theoretically.
• This distributed project can be made successful
  with the help of information technology to
  coordinate the dissemination, recollection and
  analysis of data.

7
Outline

• To keep track of the data used and produced, a
  Distributed Drug Discovery Database is generated.
• Features
• Entry of reagents in database
• Generation of combinatorial products
• Entry of synthesized products information
• Release
• Web application to enable data entry/searching of
  D3 database
• Workflow environment to support pipelining of
  tasks involving D3 database
• Web services to facilitate integration with other
  resources (e.g. PubChem)

8
Database

• Database PostgreSQL
• User Interface Java and HTML
• Tomcat Server and JDBC driver
• Chemaxon
• Marvin Sketch
• Marvin View
• JChem Reactor

9
Entity-Relationship Diagram
10
Tables

• Reagents Table
• Virtual Library Table
• Rehearsed Library Table
• Synthesized Library Table
• Source Table
• Organization Table

11
Web Interface

• Created a web interface using Tomcat 5.5, html
  and JSP.
• The front end is in HTML and the back end is in
  JSP.
• Web interface connects to the backend database
  using Tomcat 5.5.

12
WORK FLOW FOR INTERFACE
Login Page
Administrator
Student
13
Contd.
Administrator
Update the Database
Search the database
Search the Reagents
Search the products
Populating the Database
14
Enumeration Using Chemaxon Reactor

• Chemaxons Marvin Reactor enables the user to
  enumerate the virtual library.
• The Reactor program is embedded in the web
  application.
• first the user selects a reaction for the
  experiment
• from a standard set of reactions
• draw his/her own reaction in Marvin sketch
• upload using smarts.
• Second the user selects the reagents involved in
  the reaction
• The results are also available in a downloadable
  format.

15
Selecting a Reaction
16
Selecting the Reaction (contd.)
17
Setting the Reactants
18
Results
19
Student WorkBook

• integration Student Workbook into the Web
  application
• used to populate the synthesized table
• Moving data from students table to synthesized
  table is performed by the administrator
• Also used to grade students

20
Contd.

• Can be accessed at
• http//cheminfo.informatics.indiana.edu8080/ddd/f
  irst.html
• Input based on student name, team number and
  organization name.

21
Contd.

• User is given the choice of experimental vessel
  in which he wishes to input the data as shown
  below

22
Contd.

• Next the student enters the data of the
  experimental work.
• The user can enter incomplete data and then come
  back to update the experiment.
• If the user is not having sufficient data for
  input, he can input the available data and return
  back at a later time to enter the remaining data.
• The interface is capable of retrieving the
  previous information entered with the
  authentication of Student last name and team
  number.
• Marvin Sketch is embedded to enable users to draw
  the product structure and copy it as SMILES to
  enter the data in the database.
• Since the data entered could be retrieved, the
  user can always make changes if he realizes that
  he has entered incorrect data.

23
Contd.
24
(No Transcript)
25
Usability Study

• A usability study is being conducted on the
  student interface.
• To study the feasibility and usability of the Web
  application for the students.
• Points Considered for Survey
• Website design i.e., user-friendly, legibility of
  instructions, size of fields, etc.
• Use of Marvin Sketch i.e., usefulness, complexity
  and overall impression.
• Database Entry i.e., paper Vs. computer,
  updating/editing information
• Rating of the overall web application

26
Results
27
Contd.
28
Contd.
29
Contd.
30
Contd.

• Students were a bit confused in the beginning to
  use the web application.
• Though the application needs to be polished, the
  database is very useful in storing the outcomes
  of the experiments.
• Found Marvin Sketch similar to Chemdraw, but were
  confused why they were using Marvin.
• Overall the web application was a good one.

31
Administrative Interface

• Administrator is given a choice of searching the
  student input by
• Student Lastname
• Product Structure
• Team number
• Vessel number

32
(No Transcript)
33
Interface of Product Structure Input
34
Result
35
Search by Vessel Number
36
Result
37
Contd.

• Once the administrator checks the experimental
  results,
• He can grade the students
• Check for the correct product structures
• Omit the ones which are wrong and
• Input the right ones into the synthesized library
• Can send these results for further investigation
  to laboratories

38
Future Work

• Supporting more popular queries
• Extracting the Reagents from Aldrich and storing
  them in our database.
• Test and fine tuning of the distributed
  application
• Transform popular queries into web services (e.g.
  combinatorial queries into web services)

39
THANK YOU!