Questions about Registry
Should they be working on automated cloning or not?
- Reshma 14:39, 5 January 2008 (CST): The Registry should decide for themselves whether to spend their resources on automated cloning. But in general, automated cloning is a technology that synthetic biology really needs so yes!
Specifically, what are the expectations in terms of cost, turn-around time, etc.
- Reshma 14:39, 5 January 2008 (CST): Max 2 week turnaround time. Cost should be max of $50-$100. Obviously, everyone will want faster turnaround times and cheaper costs but these are my estimates for the point at which I would start outsourcing assemblies. Combinatorial libraries would require cheaper costs to be routine.
- Jason R. Kelly 23:52, 7 January 2008 (CST):$150 / 2 week per assembly
- Jeffrey Dietrich 8 January 2009: I agree with the $50-$100 price point (at least at this time) I think most people who do a lot of cloning are going to expect an average 1-1.5 week turnaround. Beyond that most of us can do it faster, and probably cheaper as well.
- reid 13:43, 8 January 2008 (CST):This is for a single assembly? I'd say two weeks is the maximum I'd want to wait. It would be nice if the delay scaled better than linear for sequential assemblies too.
Does everyone think this should be done through a synthesis company instead?
- Reshma 14:39, 5 January 2008 (CST): Since none of the synthesis companies are interested, this is a moot point in my mind.
Price-Point and Turn Around Time for a third party doing an assembly for you, by any means
- Reshma 14:39, 5 January 2008 (CST): See above.
What's the cloning amount (and type) per lab per week ?
- Seems to come in spurts, when doing dedicated construction period we do about 8-10 assemblies / person-week. Over the course of PhD probably 1-3 assemblies / person-week. Our method is standard biobrick assembly.
- Reshma 10:55, 27 November 2007 (CST): During low periods, I typically do 3-5 assemblies per week and in high periods I do 10-20 assemblies per week.
- reid 17:02, 30 November 2007 (CST) When I'm cloning, I've created ~10 constructs / week (really ~40 / month in a single batch). A construct is typically 2 or 3 domains simulateneously ligated into a backbone. I use the combinatorial / type IIs restriction system that Wendell presented at the last SynBERC meeting.
- Jeffrey Dietrich 8 January 2008 There is a pretty wide range in the Keasling lab in terms of amount cloning (and even techniques). Some people are still using the standard cloning vectors/restriction enzymes, some are now using SLIC, and some are using biobricks. However, as we get more vectors biobricked we expect that most of the lab will be using a biobrick assembly. I am doing a pretty steady 5 assemblies per week, and I think that a similar number is done for another 10 people in our lab.
What experiment can't we do that automated cloning would allow us to do?
- Jason R. Kelly 07:38, 29 October 2007 (CDT):For instance Dueber mentioned making combinatorial part libraries.
- Jason R. Kelly: Making loads of characterization constructs. Would be great to be able to insert all registry promoters upstream of the same reporter and test under the same conditions. Same for RBS's, terminators, etc. If automated cloning was in place we could continue to do this as new parts came in.
- Reshma 10:55, 27 November 2007 (CST): If assemblies were easier, I would probably scale up the number of combinations I try in parallel. For example, instead of rationally adjusting 1 parameter at a time (where a parameter is something like RBS strength, promoter variant etc.) ... I would try 5-10 RBS strength x 5-10 promoter variants x 3 different plasmid copy numbers x 3-5 different reporters. Easier assembly would also probably let me make better use of synthesis. i.e. I'll synthesize 10 different versions of my protein because I can easily try those versions with all the RBS, promoter, and reporter combinations. I've dabbled in this approach before but have found that the time I invest in building all these variants doesn't warrant what I get out of it (for my particular work ... it does pay off for others).
JCAnderson 15:44, 30 November 2007 (CST)Ditto to all that. I'd say in practice we do a similar volume of assembly, but we'd do far more if the tools were better.
- reid 17:02, 30 November 2007 (CST) I'd echo what Dueber said about combinatorial libraries. I'd also be less conservative about characterizing a large set of previously uncharacterized promoters for a particular function (yeast mating or osmo response promoters that we don't already use in the lab, for example).
JEDueber 17:42, 3 December 2007 (CST) We do about the same biobrick throughput as above, but still frequently do multistep PCR for multipart cloning when possible. Many of our constructs are making homologous repeats so we have to avoid PCR with those.