Grid Drug Discovery By Drbib

DRUG DISCOVERY & GRID COMPUTING Habibah A Wahab, PhD School of Pharmaceutical Sciences, Universiti Sains Malaysia

Biosciences WG, PRAGMA 12, March 2007

Opportunity For Computer Scientist in Pharma Industries • Ranks among the fastest growing manufacturing industries. • More than 6 out of 10 workers have a bachelor’s, master’s, professional, or Ph.D. degree—twice the proportion for all industries combined. • Fifty-nine percent of all jobs are in large establishments employing more than 500 workers. • Earnings are much higher than in other manufacturing industries. • Provided 291,000 wage and salary jobs in 2004. • Nearly 60 percent of this industry’s jobs in 2004 were in establishments that employed more than 500 workers.

Biosciences WG, PRAGMA 12, March 2007

Outlook for computer scientists in pharmaceutical industries • Computer specialists such as systems analysts, biostatisticians, and computer support specialists also will be in demand as disciplines such as biology, chemistry, and electronics continue to converge and become more interdisciplinary, creating demand in rapidly emerging fields such as bioinformatics and nanotechnology. Biosciences WG, PRAGMA 12, March 2007

DRUG DISCOVERY Post-clinical Preclinical PhaseClinical PhasePhase

Discovery Phase

Disease / Target Identification

Target Validation

Lead Identification

Lead Optimisation

Big Pharma/ Outsource to Small Companies

PreClinical

Clinical Safety & Efficacy

PostClinical

Big Pharma

•Lengthy and expensive: •10 – 15 Years; USD 800 – 1 billion •Out of 5000 medicine tested, 5 reach clinical trial (Pharmaceutical Research & Manufacturer of America) •Out of 5, 1 reach market (Tufts Center of Research on Drug Development) •Delay can cost $4-5 million/day. •Multidisciplinary Efforts involving individuals, partnerships, corporations, government agencies, manufacturer, scientific institutions. Biosciences WG, PRAGMA 12, March 2007

Source of Drugs • Drugs are derived from the following two main sources: –

Natural Product: • From plant parts or products. Seeds, stem, roots, leaves, resin, and other parts yield these drugs. Examples, salicylic acid, digitalis, vincristine and opium. • From animal sources: Glandular products from animals are used, such as insulin and thyroid. • Marine organisms: • Microorganisms: Penicillin, erythromycin • Mineral sources: Some drugs are prepared from minerals, for example, potassium, chloride, and lithium carbonate (an antipsychotic).

–

Synthetic sources: • Frequently this can eliminate side effects and increase the potency of the drug. Examples include barbiturates, sulfonamides, and aspirin.

Biosciences WG, PRAGMA 12, March 2007

Virtual Screening to Find Hit/Lead

Biosciences WG, PRAGMA 12, March 2007

Virtual Screening • Usually employ docking computation as it gives detailed representation of binding site

Biosciences WG, PRAGMA 12, March 2007

Phases of a pharmaceutical development Molecular Docking: Predict how small molecules, such as substrates or drug candidates, bind to a receptor of known 3D structure Target discovery Target Identification

Lead discovery

Target Validation

Database filtering Similarity analysis vHTS

Lead Identification

Alignment Biophores

Lead Optimization

Pre- & Clinical Phases (I-III)

QSAR ADMET

diversity Combinatorial de novo selection libraries design

Biosciences WG, PRAGMA 12, March 2007

Computer Aided Drug Design (CADD)

NaPIMM© Natural Product Information and Molecular Modelling (NaPIMM) Portal

Biosciences WG, PRAGMA 12, March 2007

INTRODUCING NAPIMm© • COMPUTATIONAL PLATFORM – – – –

Grid enabled molecular modelling platform Automatic docking & reverse docking server MD simulation server Protein-structure prediction server

• DATABASE SYSTEM: NADI© – – – –

NADI-CHEM© MNATCHEM© NADI-RA© NADI-HERBS© Biosciences WG, PRAGMA 12, March 2007

NaDI©: Features • COPYRIGHTED DATABASE SYSTEM • Comprehensive 3D chemical structure Database of natural products – Inclusive of chemical properties and biological activities references – 2D structures – 3D structures ready for in silico drug screening

• Drug Receptor Database – – provide drug receptor and disease information. – 3D structure of drug target

• Herbal Monographs – Comprehensive description about traditional uses of Malaysian and Chinese medicinal plants Biosciences WG, PRAGMA 12, March 2007

Biosciences WG, PRAGMA 12, March 2007

WHY NaPIMM ? • Malaysia has 12,000 flowering species plants (10% already known for medicinal values). • Analysis of compounds isolated from natural products showed that most of them are not available in chemical database. • Research experience in molecular modelling of natural products on : – Xanthine oxidase – protein tyrosine kinase receptor. – Dihydrofolate reductase, etc. • DOCK AROUND THE CLOCK Biosciences WG, PRAGMA 12, March 2007 •

DRUG DISCOVERY USING NAPIMM©

Database Design (Receptor & Small Molecule

Bioinformatics

Database Content Development (Receptor & Small Molecule

in silico studies

Molecular Modelling

(Physical (non-real/simplified) system that mimics biological/chemical system)

Biosciences WG, PRAGMA 12, March 2007

Enzyme Inhibition (in vitro assay)

Validation

Docking Workflow in NAPIMm literature (Target & Ligand)

2 D Structure (ligand)

3 D Structure (Target & Ligand)

Optimised 3 D Structure & Partial charge assignment

In NADI DATABASES

Ranking

Grid & Docking

AutoDock GPF & DPF setup

AUTOMATED, GENERATED WITHIN MMSERVER

Biosciences WG, PRAGMA 12, March 2007

Source of datasets • Available database; usually known chemicals – Commercial, Daylight, Cambridge CSD, MDL etc – Public domain e.g. ZINC, NCI, PubChem

• Totally virtual compounds generated by computers (either has existed or yet to be existed). • Or build one yourself; perhaps a dedicated one like natural product database Biosciences WG, PRAGMA 12, March 2007

Natural Products: Facts Plants

Extracts

Fractions

Chemicals

H3C O

H3C

H3C

H2N H3C

H3C

CH3

O

OH

OAc

CH3

O Norcarotenoid

CH3

O Jaspine B

OSO

OH OH

HN

O

H HN 3

NH

NH

2

2

H

CH3

O Jaspine A

Biosciences WG, PRAGMA 12, March 2007

squalamine

OH

3

Natural Products: Drug Source

Biosciences WG, PRAGMA 12, March 2007

Natural Products: Drug Source • 65% drugs exist today originally derived from natural sources. • E.g. digitalis, aspirin and paclitaxel (Taxol) • “natural products are biologically validated starting points for library design”. • Synthesised/modified by proteins – highly likely to bind to proteins having similar folds.

Biosciences WG, PRAGMA 12, March 2007

Biosciences WG, PRAGMA 12, March 2007

Cluster D ss

Ha

Vi ew

User

wk ,

or

lo g

in

&

do

up

wn l

lo

oa

d

re

ad

su

lt

lig

an

d

in p

ut

Output files generated and uploaded to Hawk

ce

Receive input files from Hawk and run docking jobs on each assigned nodes

Ac

Receive input files from Hawk and run jobs on each assigned nodes Cluster C

Output files generated and un r uploaded to Hawk d es n a od wk d n a H ne m ig ro ass f nd s a h e fil ac ed at wk ut n e r p in s o ne Ha e e b g v i o to ce ng j les ed i e i f R ck d t pu ploa t do Ou u

p

ss

ce Ac

file

d loa

H

k, aw

&u n i log

w

Vie

d loa

file

n

di

an lig

t pu

ult

User

res

n ow

d or

Submit input files and distribute jobs to several clusters on Grid environment Process ligand input file, create parameter Hawk server – Malaysia files & shell script files O ut Re (hawk.usm.my) pu ru ce t u n pl file do ive oa s ck inp de ge Input files in u d ne g tf to ra jo ile bs s H ted aw no on fro k and de e m s ac Ha h as wk sig an ne d d

Grid environment Cluster B

Cluster A Biosciences WG, PRAGMA 12, March 2007

NaPIMM CASE STUDY Screening of flavonoids on Dengue-2 Protease

Biosciences WG, PRAGMA 12, March 2007

Dengue • Dengue is a serious infectious disease that is endemic in over 100 countries. • An estimated of 50 million infection per year globally with more than 2.5 billion people are at risk for epidemic transmission. • Major affected regions are the south-east Asia and the western Pacific (increasing reports of this disease in the Americas). • Caused by the dengue virus which is a member of the Flaviviridae. • Spread by the highly domesticated Aedes aegypti mosquito. • Dengue can be classified into: – Dengue Fever (DF) •

a flu-like illness with symptoms like fever, headaches, joint aches and rashes

– Dengue Haemorrhagic Fever (DHF). • more severe and often fatal complication of DF as a result of the dengue shock syndrome (DSS).

• To date, no licensed vaccine or therapeutic drug available for DF and DHF/DSS, although there have been reports of some vaccine candidates in clinical trials. • The treatment for DF and DHF/DSS has only been supportive thus far. Biosciences WG, PRAGMA 12, March 2007

Flavonoids • Flavonoids belong to a class of chemical compounds that are ubiquitous in the vegetable kingdom with about 6000 structures of flavonoids are known.

OH OH O

• Among the most widely studied flavonoids HO is quercetin (Fig. 1) which is present widely in vegetables and fruits, with a daily intake of up to 25 mg/day in a normal human diet. • Most flavonoids are known to be a powerful antioxidant that can scavenge the reactive oxidative species and a potent chemopreventive agent. • Now, we would like to see whether it has anti-dengue activity

OH OH

O

The chemical structure of quercetin

Biosciences WG, PRAGMA 12, March 2007

Screening of flavonoids on Dengue-2 Protease

Biosciences WG, PRAGMA 12, March 2007

Screening of flavonoids on Dengue-2 Protease

Biosciences WG, PRAGMA 12, March 2007

Screening of flavonoids on Dengue-2 Protease B. rotunda (1) B. rotunda (2) B. rotunda (3)

Biosciences WG, PRAGMA 12, March 2007

WHAT’S NEXT • We are very interested to further develop AMEXg, MMServer on GridSphere • Further develop NAPIMm so that it fully utilise Grid Computing – current test only involve 4 PRAGMA sites. • To incorporate Quantum Chemical Calculation and Cheminformatics application within NAPIMm • Use NAPIMm for PRAGMA real science project • Have PRIUS students in USM to help us achieving above.

Biosciences WG, PRAGMA 12, March 2007

Sipadan Island, Sabah, Malaysia

I would like to thank all system administrators within PRAGMA community especially: Nguyen Viet Han dr Vietnam, IOIT Eduardo Murrieta León, Mexico, malicia cluster Yusuke Tanimura, Japan, AISTX cluster

Biosciences WG, PRAGMA 12, March 2007

Acknowledgement Academic Collaborators      

     

Dr. Nornisah Mohamed Dr. Rashidah Abdul Rahim Professor Noorsaadah Abdul Rahman Professor Nazalan Najimudin Dr. Razip Samian Dr. Chan Huah Yong Professor Janez Mavri Professor Tom Scior Dr. Stephen Doughty Dr. Andre Aubry Dr. Regis Venderesse Dr. Bridget Jamart

Postgraduates:  

       

Belal O Najjar (MSc Candidate) Nur Hanani Che Mat Choong Yee Siew Choy Sy Bing Ezatul Ezleen Kamarulzaman Imtiaz Khalid Ismail Mohd Shah Nurul Bahiyyah Ahmad Khairudin Yam Wai Keat Suriyati Muhamad

Biosciences WG, PRAGMA 12, March 2007