cheminformatics workflows using mobile apps for drug discovery
DESCRIPTION
talk for SLAS 2014 slidesTRANSCRIPT
Alex M. Clark1, Antony J. Williams2 and Sean Ekins3,4
1 Molecular Materials Informatics, 2 Royal Society of Chemistry, 3Collaborations in
Chemistry, 4 Collaborative Drug Discovery, Inc.,
Cheminformatics Workflows Using Mobile
Apps for Drug Discovery
• mobile is revolutionary: a clean break
entirely new user interface
no backward compatibility
highly constrained resources
applicable to entirely new situations
mainframes
minicomputers
personal computers
portable laptops
mobile tablets
smartphones
?
The Computing Revolution #3
Williams et al DDT
16:928-939, 2011
Arnold and Ekins, PharmacoEconomics 28: 1-5, 2010
Williams et al., In collaborative computational technologies for
biomedical research 2011
Fitting Mobile Apps into R&D Workflow
Chemistry Apps • Reference data
• Education
• Structure drawing
• Database searching
• 3D viewing
• Reactions & collections
• Property calculation
• Model building
• Graphical presentation
• Data sharing
http://goo.gl/Goa4e
Need for dedicated website / store for science Apps
– find out more at www.scimobileapps.com
Simple App Workflows
• Look up structure in ChemSpider
• Saving structure as molfile - open in MMDS
• Run substructure search in ChEBI using MMDS webservice
• Open molecule from MMDS and assign scaffolds in SAR Table
Generate substituents
• Predict missing activities for compounds in SAR Table
• Suggest compounds to make in SAR Table
• Find a reaction in SPRESImobile
• Use Yield101 to calculate synthesis yield
• Share data with Dropbox using MolSync app
• Tweet a reaction with MolSync
• Read the data with ODDT mobile app
Clark AM, Williams AJ and Ekins S, Chem-Bio Informatics Journal, 13: 1-18 2013.
A Bigger Vision
• Mobile chemistry originally intended to support
desktop workflows
• Mobile+Cloud can be a total replacement
• Entirely new user expectations for apps:
- easy to learn
- delightful to use
- trivial to install
- inexpensive or FREE
• Extremely disruptive to existing software vendors!
APPIFYING DATA - From PDF to Mobile App
Lots of data
but how to
make it useful
for chemists?
Chemists see
structures
PDF not
accessible,
small text- too
much data
http://bit.ly/GzQ5ty
Green Solvents and Lab Solvents FREE Apps
Alex Clark made the App in 3 days
Android version – Lab Solvents
Includes GSK solvent data
ACS Sustain Chem Eng 1: 8-13 (2013)
TB Kills 1.6-1.7m/yr (~1 every 8 seconds)
1/3rd of worlds population infected!!!!
Multi drug resistance in 4.3% of cases
Extensively drug resistant increasing incidence
No new drugs in over 40 yrs until Bedaquiline
Drug-drug interactions and Co-morbidity with HIV
Increase in HTS phenotypic screening
1000’s of hits no idea of target Use of computational methods with TB is rare
30 years with little TB mouse in vivo data MIND THE TB GAP
IN V
IVO
INA
CT
IVE
IN V
IVO
AC
TIV
E
Pathway analysis
Binding site similarity to Mtb proteins
Docking
Bayesian Models - ligand similarity
Predicting the target/s for small molecules
TB molecules and target information database connects
molecule, gene, pathway and literature for >700 molecules
iPhone Android
TB Mobile layout on iPhone and Android
TB Mobile Molecule Detail and Links
iPhone Android
Molecules active against Mtb evaluated in TB Mobile app
Workflow from sketching molecules in MMDS mobile app
to exporting and opening with TB Mobile
http://goo.gl/UTTH0
TB Mobile – poster on Jan 2013
http://goo.gl/vPOKS http://goo.gl/iDJFR
TB Mobile – Is on iTunes and Google play
and it is FREE
http://goo.gl/7fGFW
What next ?
Update with more data
Add a weighting or scoring function to account
for heavily populated targets
Expand beyond the similarity measure
Add algorithms to predict activity
Could we appify data for other diseases/ targets
Connecting
data/tools
like a TB Spider
In vitro data
In vivo data
Target data
ADME/Tox data
& models
Drug-like scaffold creation
TB prediction tools TB publications
• Preliminary work done with desktop software: com.mmi
• Fragment TB Mobile structures, scaffold-like
• Perform scaffold-substructure vs. 7000 in vitro
• Derive R-groups, tidy, present graphically, browse...
TB Mobile in a TB Workflow
Source Materials • Scaffold:
• Scaffold origin: inhibitor of Glf target
• 87 molecules with in vitro activity (yes/no)
• Scaffold seems to elicit an activity pattern
• Next step: load it into the app ecosystem...
http://molmatinf.com/
venice.html
To see the rest
of the TB
workflow……
Open Drug Discovery Teams
• Curation of open data, e.g. Twitter & RSS feeds
• Rare & neglected diseases, precompetitive
areas
Harvested Tweet
• Tweet got harvested into Tuberculosis topic
• Inline preview browsed, with other
thumbnails
What we can do now… Take HTS screening hits Query public databases Propose targets Design / purchase analogs Predict activity All on a mobile device / anywhere
Conclusion
• Cheminformatics workflows historically the role of
specialists: expensive and/or complex
• Mobile apps are much cheaper and much more
accessible to experimentalists
• Mobile+cloud can:
- replace simple-to-medium tasks
- coexist with complex tasks run on desktop software
• Other advantages:
- anywhere/anytime portability
- excellent collaboration and sharing
- non-existent installation or maintenance burden
27
Malabika Sarker, Carolyn Talcott, Joel Freundlich, Barry Bunin
2R42AI088893-02 “from the National Institute of Allergy And
Infectious Diseases. (PI: S. Ekins)
You can find me @. CDD Booth 653..
Poster 224
PAPER TITLE: “Dual Response and dataset Fusion for
Machine Learning Models for Hit to lead Optimization in
Mycobacterium Tuberculosis Drug Discovery”
Monday, January 20, 2014
Presentation Time: 1:00 PM – 3:00 PM
Acknowledgments