Your SlideShare is downloading. ×
0
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Meta QSAR
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Meta QSAR

1,015

Published on

Presentation on using the Discovery Bus to develop a new field of research, "Meta QSAR" the comparative study of QSAR modelling methodology. Given at UK QSAR Society meeting at Syngenta October 22nd …

Presentation on using the Discovery Bus to develop a new field of research, "Meta QSAR" the comparative study of QSAR modelling methodology. Given at UK QSAR Society meeting at Syngenta October 22nd 2009

Published in: Health & Medicine, Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
1,015
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
24
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. Meta QSAR David Leahy
  • 2. Structures Data Descriptor Descriptor Descriptor Descriptor Test Set Filter Filter Filter Filter Model Model Model Model Transform Transform Train Train Predict Predict Predict Predict Predict Predict Predict Predict
  • 3. Calculate Descriptors Filter feature request ... Model build request ... Calculate descriptors request ... responses responses responses Planner Model Build Filter Features
  • 4. Calculate Descriptors Filter feature request ... Model build request ... Calculate descriptors request ... responses responses responses responses responses responses Planner Model Build Filter Features Calculate Descriptors
  • 5.  
  • 6. Meta QSAR. Methods <ul><li>Wombat Database (thanks to Tudor) </li><ul><li>First 80 data sets (human, n &gt;75, Range &gt; 2 Logs)
  • 7. Miscellaneous (HSA, Herg, CHI, Cl I ) </li></ul><li>Descriptors </li><ul><li>CDK, CDL, MOPAC, HQSAR, E-State, H-State, LSER, AlogP, XlogP </li></ul><li>Filter Feature </li><ul><li>Hall method </li></ul><li>Model Building (Continuous) </li><ul><li>RNN, Rlinear, Rpart, RPLS, GUIDE </li></ul></ul>
  • 8. &nbsp;
  • 9. Model Assessment &amp; Selection <ul><li>Validity </li><ul><li>Statistical tests
  • 10. y-scrambling </li></ul><li>Stability
  • 11. Domain of Application </li></ul>
  • 12. &nbsp;
  • 13. &nbsp;
  • 14. &nbsp;
  • 15. &nbsp;
  • 16. &nbsp;
  • 17. &nbsp;
  • 18. &nbsp;
  • 19. Meta QSAR: Automation of Decision Making in QSAR Modelling
  • 20. Mother of All QSAR Study <ul><li>Expand data series </li><ul><li>Wombat Database (~ 1000 datasets)
  • 21. EBI Stardrop Database (~2-3000 datasets)
  • 22. 0.5 million models, 25 years </li></ul><li>Clouds </li><ul><li>Amazon EC2
  • 23. 100 Microsoft Azure nodes </li></ul><li>New methods </li><ul><li>Simpler to add, any technology </li></ul></ul>
  • 24. Amazon Master Server Amazon Worker Servers Azure Worker Servers Security Virtual Teams Inkspot Science Sci Software Services Workflow Cloud
  • 25. Inkspot Science Agents
  • 26. Inkspot Install to Azure
  • 27. Work in Progress <ul><li>MOAQ Study </li><ul><li>Classification Model Analysis
  • 28. Domain of applicability and local models
  • 29. Temporal QSAR </li></ul><li>Open QSAR </li><ul><li>MOAQ Results on-line
  • 30. New methods install and benchmark
  • 31. New data series
  • 32. Property prediction </li></ul></ul>

×