SBPWG:Meetings/Nov 15 2011

Home        About        Members        Meetings        Gaps        Discussion        Resources       

Meeting Notes: November 15 2011

November 15th 6:30-8:30pm @ UCSF Mission Bay (room TBA)

• Chris Anderson, assistant professor of Bioengineering at UC Berkeley and SynBERC PI, joined the working group to pitch, discuss and get feedback on several prospective biosafety-related projects addressing challenges in biological design, education and management.

Attendees

• Megan Palmer (Stanford)
• Ryan Ritterson (UCSF)
• Jay Vowles (Stanford)
• Connie Eaves (UCSF)
• Veronica Zepeda (UCSF)
• Andy Chang (Stanford)
• Josh Wolf (UCSF)
• Nathan Hillson (JBEI)
• Kevin Costa (SynBERC)
• Chris Anderson (Berkeley)
• Derek Greenfield (LS9)
• Mike Fero (Teselagen)
• Eduardo Abeliuk (Teselagen)
• Ellick Chan (Stanford)

Slides

See slides here

Audio

Download audio file here

Pictures

Notes

Notes transcribed by Ellick Chan

Please see the timestamps below to jump to a particular section.

16:00

• Start
• Working Group Origins at Synberc Spring retreat. Issues covered in practices. Translation, safety, security.
• Recombinant DNA, film
• Discovery Days science festival
• Biotech bootcamp

18:30

• Megan intro to Safety issues

19:00

• Chris: topics covered
  • 1) Containment
  • 2) Training
  • 3) Assessing risk

Don't want people making smallpox - not directly issue of synthetic bio. No one is working in that area.

Politics of biosafety - economics. Models for risk. Responding to risks. No financial collapse. Actively mitigate risk.

23:00

• Making an effective containment strategy - Auxotrophy
• Self-kill device in cell: problems
• containment of org
  • codon swapping
  • containment of DNA
  • Dapd auxotropes
  • unnatural amino acid dependence
  • self destruct device
  • synthetic auxotrophes

• What's the edit distance to something dangerous?
• What is dangerous?

• Lack of a theory to assess danger

28:00

• modifications cannot escape by complementation
• cannot revert through natural mutation
• sufficiently inexpensive to use in a bioreactor
• no adverse impact on viability of the cell
• be extensible to industrially relevant organisms

31:00

• pick a gene that can't be knocked out or complemented
• small molecule dependency
• allele exchange of new copy with mutant copy
• 300 essential genes in ecoli
• tour based selection system

36:00

• another paper based on a similar technique
• politics
• design to contain nature. if it breaks out, it may be our fault.
• what's the bar for regulation?
• assay to prove safety/effectiveness
• not our role to set our own bar - someone else should set it

39:00

• ques: academic research on rates? chris - industry will say so

41:00

• training
• NSF needs curriculum for SynBERC
• Berkeley has classes, but most other institutions don't
• Ethics class?
• Online training module for biosafety
• heart of the problem - no theory on safety
• BUA - Biological Use Authorization -> NIH guidelines
• explosion of genes - no one knows all the risks of them off the top of their head

49:00

• risk assessment
• 6 categories
  • virulence
  • environmental stability
  • communicability
  • quantity
  • vaccine
  • allergenicity
  • potential for nefarious use

• sign off by committee if necessary
• never seen them say no. may give you recommendations
• rarely/never re-review in the case of new identified risk

54:00

• quizzes are basically a checklist of thoughts that the student should consider
• megan - does such a flowchart exist?
• biosafety guys aren't that scared of synth bio compared to other risks
• subjective questions
• risk should be considered from factors other than the originating organism

64:00

• assessment
• risk based on features
• if it can be programmed, there should be enough formalization to do so
• formalism simplifies things
• evaluation of risk depends on biologist's perspective
• first attempt

70:00

• minimize risk - example in radiation safety
• payload delivery
• risk group is max risk at any stage
• transposon not caught
• chassis can be dangerous
• device can be more dangerous than sum of its components
• substance could be dangerous in a different context

80:00

• unintentional promoter can express gene in unknown ways
• accidental recombination
• act ontology
  • feature, family, device
• autorepression device

88:00

• verification and synthesis
• families - by hand
• features - partially automated
• main issue: how to populate list
• all knowledge in literature, but not enough manpower to pursue
• inference of devices resulting from modifications

92:00

• map dangerous functionality to devices
• scenarios of things that are dangerous, but not obviously dangerous
• engineered peanut allergy on salmonella
• curate high level families for danger
• what's the difference between a strain (cell+genome+plasmid mods) and a device(modification, composition of devices can create a strain)?
• strains as combination of devices. complex network.
• look at interactions between added features + original genome

100:00

• Peanut peptide combining with transporter
• Megan: is there a theory on what constitutes danger on a given agent. Weaponization of anthrax - what's the special knowledge needed to make it harmful
• 2001 list doesn't consider starting path and possible paths to dangerous agents
• megan: can we classify existing dangerous agents and process of creation?
• megan: how do we hedge our bets in making sure we're doing good work in this area? From DARPA's perspective? Safe chassis development vs automated monitoring vs surveillance vs forensics
• megan: partnerships/funding
• nsf cares about training
• containment
• megan: orthogonal genome vs selective growth environment

112:00

• orthogonal genomes very academic, but probably not industrializable - whole platform required to make this work
• what are axes of risk?
• danger from synergistic reaction adding to risk
• expert system, review committee

120:00

• megan: is there a politically-safe way to get a group to look at these issues? what would pr think?
• most of these are obvious
• is there an arms race for risk assessment schemes?
• ACT theory is open knowledge
• value will be in list of devices to query for risk
• this can be used to find dangerous devices…

Wrap-up:

• Naming devices and rubric is interesting
• Doing exercise around context of training module
• Siebel foundation event coming up

• New ways to fund biosafety
• OSTP request for info - december 6th
• Leaders in practice, 1 week bootcamp in summer - in DC