Open worm public hangout 10-08-11


Published on

Introduction to the Open Worm project, progress for release 1, plans for release 2.

Published in: Technology
  • Be the first to comment

  • Be the first to like this

Open worm public hangout 10-08-11

  1. 2. What I cannot create I do not understand Richard Feynman’s last board
  2. 3. A multi-scale data problem Scale Whole brain data (20 um microscopic MRI) Mosiac LM images (1 GB+) Conventional LM images Individual cell morphologies EM volumes & reconstructions Solved molecular structures
  3. 4. Reverse-engineering? A system whose mechanisms are obscured
  4. 5. What is reverse-engineering? System whose mechanisms are obscured Individual components and an explanation of how they fit together
  5. 6. What is reverse-engineering? Provide a framework of parts ready to be snapped together
  6. 7. Put the parts back together
  7. 8. A multi-scale synthesis problem
  8. 9. Simulation-based research <ul><li>Challenging intuition via predictive models </li></ul><ul><li>Construction of models requires deep organization of understanding of a system </li></ul><ul><li>Forces you to fill in gaps </li></ul><ul><li>Observing unexpected behaviors produced by the models </li></ul><ul><li>Able to look across multiple variables at the same time </li></ul><ul><li>Can do experiments that would otherwise be impossible to do in a real biological system </li></ul>
  9. 10. Enter the worm: c. elegans What’s up, baby?
  10. 11. Virtual physical organisms in a computer simulation
  11. 12. In search of nature’s design principles via simulation <ul><li>How can a humble worm regulate itself? </li></ul><ul><ul><li>Reproduces </li></ul></ul><ul><ul><li>Avoids predators </li></ul></ul><ul><ul><li>Survives in different chemical and temperature environments </li></ul></ul><ul><ul><li>Seeks and finds food sources in an ever changing landscape </li></ul></ul><ul><ul><li>Distributes nutrients across its own cells </li></ul></ul><ul><ul><li>Manages waste and eliminates it </li></ul></ul><ul><li>If we can’t understand genes to behavior here, why would we expect to understand it anywhere? </li></ul>
  12. 13. Enter the worm: c. elegans I’ve only got 1000 cells in my whole body… please simulate me!
  13. 14. A complete simulation of the worm’s brain, body and environment Simulated World Detailed simulation of worm body Detailed simulation of cellular activity
  14. 15. The goal: understanding a faithfully simulated organism end to end Extracting mathematical principles from biological systems is necessary if we are going to understand and reconstruct the much larger system of the human.
  15. 16. Outreach: put the model online and let the world play with it <ul><li>Sex: Hermaphrodite </li></ul><ul><li>Interested in: Escaping my worm Matrix </li></ul><ul><li>Relationship status: Its complicated. </li></ul>
  16. 17. Worm biology <ul><li>~1000 cells / 95 muscles </li></ul><ul><li>Neuroscience: </li></ul><ul><ul><li>302 neurons </li></ul></ul><ul><ul><li>15k synapses </li></ul></ul><ul><li>Shares cellular and molecular structures with higher organisms </li></ul><ul><ul><li>Membrane bound organelles; </li></ul></ul><ul><ul><li>DNA complexed into chromatin and organized into discreet chromosomes </li></ul></ul><ul><ul><li>Control pathways </li></ul></ul><ul><li>Genome size: (9.7 x 10 7  base pairs or 97 Megabases), vs human: 3 billion base pairs (3 X 10 9  bp or 3000 Megabases). </li></ul><ul><li>C. elegans homologues identified for 60-80% of human genes (Kaletta & Hangartner, 2006) </li></ul>
  17. 18. C. Elegans disease models Kaletta & Hengartner, 2006
  18. 19. Can present drugs Kaletta & Hengartner, 2006
  19. 20. Entire cell lineage mapped
  20. 21. Entire cell lineage mapped
  21. 22. Entire cell lineage mapped
  22. 23. Entire cell lineage mapped
  23. 24. Full connectome Varshney, Chen, Paniaqua, Hall and Chklovskii, 2011
  24. 25. P. Sauvage et al. / Journal of Biomechanics 2011 Biomechanics
  25. 27. Interrogation of Behavior
  26. 28. Core platform: Open Worm project
  27. 29. One core hooks together multiple simulation engines addressing diverse biological behavior
  28. 30. Estimates of computational complexity <ul><li>Mechanical model </li></ul><ul><ul><li>~5 Tflops </li></ul></ul><ul><li>Muscle / Neuronal conductance model </li></ul><ul><ul><li>~240 Gflops </li></ul></ul><ul><li>One Amazon GPU cluster provides 2 Tflops </li></ul>Source:
  29. 31. Mechanical model Palayanov, Khayrulin, Dibert (submitted)
  30. 32. 3D body plan Christian Grove, Wormbase
  31. 33. Team – A brief history
  32. 34. Collaboration technologies used
  33. 35. Jan – Sept 2011
  34. 36. Architecture
  35. 37. Neuronal model GPU Performance Testing: 302 Hodgkin-Huxley neurons for 140 ms (dt = 0.01ms) Architecture proof of concept using Hodgkin-Huxley neurons ms
  36. 38. Worm Browser _0
  37. 39. Physics: SPH Smoothed particle hydrodynamics (SPH) algorithm for soft-body / liquid finite element interactions
  38. 40. Soft-body & fluid mechanics
  39. 41. Finite element modeling
  40. 42. C.elegans neuron models in NeuroML
  41. 43. Mendeley group has 234 references
  42. 44. Presented poster at Neuroinformatics 2011
  43. 45. OpenWorm links: <ul><li>Project page: http :// / </li></ul><ul><li>Twitter: @openworm </li></ul><ul><li>Mailing List: openworm </li></ul>
  44. 46. <ul><li>Sept 2011 – March 2012 </li></ul>
  45. 48. Muscle cell with “arms” Cell Body 5 arms, 10 compartments each, passive currents Cell body, 1 compartment, active currents Boyle & Cohen, 2007
  46. 49. Case study: locomotion Gao et al, 2011
  47. 50. Conductance model of c. elegans muscle cell Boyle & Cohen, 2007
  48. 51. Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Cell Body Quadrant 1 Quadrant 2 Quadrants of muscle cells
  49. 52. Genetic Algorithms and Parameter optimization Achard, De Schutter, 2006
  50. 53. Gaming and crowdfunding