[DIYbio] Re: A better way to share/use DNA?

Josiah, thank you for taking the time to give such thorough feedback. I'd also like to apologize if some of our core concepts were unclear to you, so I'd like to take the time to clarify those here.

To provide you with some background, we started building AutoClone with commercial research teams in mind, especially those working in high throughput labs with tens of thousands of sequences under management. While scientists generally know what they work with/what they have, on this kind of scale it often boils down to a numbers game - repositories this large cannot be memorized or easily re-used so this information has to be put on a computer in order to be made accessible. This problem is compounded as scientists leave and join the company over many years – people just forget what's in there or simply don't know.

While 10 years worth of plasmids and primers looks disgusting (and I have personally seen and experienced this in academic labs), I'd point out that the companies we're talking about have "working" freezers and "stock" freezers – the former being day-to-day work by staff and the latter being tidy, well curated and accessible. In fact, most of the time department leads are excited by the prospect of "unlocking" the value in those freezers in new ways.

Indeed, there is significant value in there, especially when the DNA and corresponding sequence data within can be arranged and stored in part/feature format. The idea of organizing this information in a manner that makes it easy to extract and re-use parts, or try novel combinations of parts, has gained a lot of traction with the companies we are working with.

With respect to an "assembly solution", we aren't talking about a tool that finds 500 10mers and tells you how to stitch those together to form a 5kb vector of interest. Rather, we are talking about careful automated analysis of an inventory, and identification of whole fragments, promoters, terminators, expression cassettes and so on. We actively filter out "micro-PCRs" and deliver the solution that uses the fewest separate parts, the fewest different sources, and calculates the shortest length of DNA you'd need to synthesise. The solutions are computed across a range of cloning chemistries including cut and paste, and homologous recombination strategies. Many different solutions are possible, and the user is free to pick the one they like the most. Eventually we would like to add the ability for the user to modify and combine solutions. 

Third party repositories have been preloaded into AutoClone. So if there is a commercially available vector from Promega that has a fragment you need, we will tell you that and design primers to amplify it and bind it to another fragment that is identified. If the fragment is available from Addgene, we will tell you that. If its available locally, we can prioritise that. Instead of you having to independently test if what you need exists somewhere, we automatically find it for you and let you know how to take advantage of this to save a ton of time and money.

I would say this goes far beyond Addgene's service. I've not heard about an excel sheet that does this :P

I personally spend most of my time in the lab testing our software and have had plenty of experience designing and ordering untold primer sets and also in trying to get plasmids from other people.  I agree that entering primer data into AutoClone over and over again would be a horrible experience, and that the general lack of time for people to do any data entry is an issue. Thats why its better to avoid it altogether.

We are looking to solve this by partnering with synthesis companies such that users can directly order their primers through the tool, and as such, have the primers automatically entered into their inventory in this manner. This method would create a type of "one-stop shop" for ordering and managing your oligos in addition to your plasmids and constructs; this means a lot less work for you. 

Did the primer work? Great! Put a note in it. That's what you'd do in your lab book, so why not log that in your account, and choose to tell other people if you want. For companies, simply being able to look back at their sequences and see exactly how they were assembled from other sequences (and where they came from), and using which primers – all that data is useful to track, especially for bioproduction teams that often have to provide customers with precise info on how things were made. 

You asked if I have ever asked someone for a plasmid. I have, but to be honest I don't like doing it.  We are fully aware of the trust issues surrounding plasmid usage. Personally, I would never feel comfortable using a plasmid that came from an academic that I have never worked with before. For this reason, one of our top priorities is developing a solution that allows you to view validation data associated with these parts. Id also note that within a commercial setting, researchers working for the same entity tend to trust each other more as things are done to a shared standard. This has reduced the impact of the sharing-trust problem for us.

Nonetheless, in order to prevent academic scoops and create a measure of privacy for researchers, we have made it so that sequences can be selectively hidden from everyone in terms of viewing it. Additionally, the researcher has the ability to choose whether or not to the let AutoClone have certain sequences "count" towards a solution on anyone else's account.

Again, I'd like to note here that this is not something devised specifically for academic labs. Rather, it was geared towards companies that have a need to organize and re-use limited resources, reduce overheads, and build things collaboratively without sending mass emails between departments.

There is some overlap between problems in commercial settings and the DIYbio community. The four part problem specific to the DIYbio community is: the lack of connectivity between groups, the lack of synthesis options, the lack of a system that fosters the sharing of parts, and the difficulty of organizing and communicating outside of single hackspace. 

I am bringing AutoClone to your attention because I see it as a good way to combine all the repositories of every hackspace within this tool, and then let every hackspaces have insight and access to each other's inventories. Because biohackers are on a tight budget (oftentimes even tighter than academics), and given the relative lack of synthesis support for biohackers by the big synthesis companies, it would make sense to share DNA and discuss projects across a platform.

For instance, if one hackspace makes the investment to buy in or build a vector, I believe the community should have access to it, and I'd like to help make this happen if I can. We have already seen Koeng's kind offer to make everything he has available to anyone else – I don't think he is unique here. The DNA that each biohacker would have access to would expand massively with our tool, and we'd love to help make this happen.

We are working with an increasing number of iGEM teams using the software to curate their parts, share parts with each other, and search many thousands of commercially available vectors for sequences that could be useful to their work. The teams we are collaborating with aim to pass these accounts down to next year's teams (previously, teams at the same university have tended to not have access to the notes or parts of previous years as there is no tool to log this). I'd like to arrange for these iGEM teams collections to be made available to biohackers and see where it takes us all.

Eventually what I would like to do is find a way to use the tool to quickly create variants of designs (from a library) that could be distributed amongst the community for different teams to build and compare. I.e if there is a very cool project that the community coalesces around and wants to build and test different variants of, individual teams could each take on a few designs that make use of what they have, while others do the same for their inventories. This is some while off though.

I fully recognize that my position within the company may warrant some healthy skepticism (and I welcome it) with respect to my viewpoints and analysis. But aside from this identity, I also see myself as a researcher who has been living with this tool and working to improve it since it's inception. I think we've built something which has the potential to streamline and organize DNA management and construction, and also address some of the wider needs of the DIYbio community. We are always looking for ways to make it better and certainly aren't looking to charge anyone for it out of nowhere.

Thanks again for taking the time to give such thorough feedback. I am looking forward to continuing this exchange. If anyone else has any thoughts on any of the above points, please jump in.

If people are working in a lab together they most likely are either working on the project together so know the exact sequences that are ordered or are working on completely different projects so as to not compete with each other and so requesting stuff from each other is probably not going to happen except in a few really huge labs. Or will happen so rarely that creating a huge database of everything is not worth it. Have you ever seen 10 years of plasmids and primers? We are talking tens of thousands of tubes in even small-medium size labs and probably hundreds of thousands in big labs. A simple website is not going to fix the organization of this.

Is this just a distributed DNA repository instead of a plasmid repository? The amount of time it takes to run any significant DNA or plasmid sharing repository is hindrance to most labs. It's not that people don't want to share DNA it is that it requires work to do so. Why do you think there is no cheap commercial repository for sharing plasmids? This is why addgene can charge $60 / plasmid. I have thousands of sequences and if someone wants a set of primers I have I am not going to send them to someone I don't know when they cost $20 or less to order on IDT. By the time someone in a lab takes the time to find them aliquot them and mail them you will have definitely received your DNA from IDT. It is fairly obvious that you probably have never asked anyone for a plasmid or DNA before.

Assembly solutions what does that even mean? Assembling DNA from random fragments is not something anyone wants to do and the possibilities and probabilities of being able to do this are extremely small especially if only using popular restriction enzymes. Is your program going to say that the gene is available? Sounds like a basic search function... Sounds like addgene.

The problem with keeping an inventory is not that there is no place to keep an inventory. Excel can keep an inventory of primers and plasmids and there are probably tens of websites out there that also already keep an inventory if people want. Have you ever ordered 50 sets of primers? or 100?  or 1000? Who is going to enter all these into the database and organize them? That is the bottleneck not needing a database.

Most DNA sharing between academic institutions requires the receiver to have a FedEx number to pay for shipping. If people on your website are paying for the DNA and replication and shipping it is easy to see how this method won't be very successful and how most won't take part.

Someone from a lab is not going to trust a 10 year old set of primers(maybe some people but I definitely would not). How do I know if they are contaminated or not? How do I even know what the person was using them for? Did the person purposely or accidentally mix primers in the tube for a primer mix? When it costs less than $20 to order new primers would you bet your experiment and sanity on 5 or 10 year old primers?

Most DNA is not reusable. It is designed for a singular purpose or purposes that are very similar. If I order a gene or primers they can't be reused by someone for a different purpose or it would be extremely difficult for that. Maybe one could perform multiple restriction digests on a gene if they are lucky(but usually restriction sites are avoided inside genes) and then a PCR and then a ligation but this becomes more expensive and much more time consuming than just ordering the DNA you want from IDT or another company. If I have DNA only as a gene that most likely means I have just started working on it. If someone else wanted this gene most people would not send it to them because that is usually direct competition. And I honestly don't know one lab who would ask someone for an uncloned gene. Very bad form in most Science circles. Order it yourself. This seems like a better idea for DIY Bio but ....
Plasmids one can miniprep a bunch of and the plasmid is relatively pure. However, when one orders a gene from IDT how do I send a sufficient quantity of the gene to someone when the company only sends me 200ng? I could PCR it but would you really trust a PCR from someone you never met and know nothing about? So that means I need my own primers and I need to sequence the product they sent me. Most genes are ordered with flanking restriction sites or custom tags which makes direct PCR without mutations or modifications unlikely. In the end the time and effort and money this requires is probably more expensive then ordering the gene for yourself in all but the most extreme cases. Especially with the prices of IDT and others dropping so low recently. You are asking for the sender and receiver to both put in some significant effort to share. Still the reusability of most DNA fragments is so small it is hard to create any sort of significant participation.

Primers that are commonly used(T7 fwd, T7 rev, M13, &c.) which could be the most requested things is like asking someone from another university if you can borrow a ream paper or a box of pipette tips. They would stare at you blankly.

Finally, Scientists are not idiots we don't reorder DNA on accident that we already have as you imply on your website("Eliminates wasted resources by avoiding the synthesis or purchase of existing DNA fragments.") at least not in any lab I have ever been in. And a number of labs I know keep an excel spreadsheet of plasmids or primers that are easily searchable....

In closing, I think the space you are attempting to enter is already full and your idea needs some work because the practicality of it all is very small. But feel free to rebut many of the claims I make. I would be interested to hear your response.



On Monday, June 23, 2014 3:30:09 AM UTC-7, EdwardPerello wrote:

If two people work together in a lab closely, then they will probably have a good idea of each other's DNA sequences/repositories. This information makes it easier for them to borrow and build on each other's work. However, if two people in a lab don't work closely together, they won't be able to do this and any improved productivity from mutual sharing transactions is lost.

My startup company, Desktop Genetics, has built a tool supporting networked DNA repositories for individual scientists/lab groups. Each person or group uploads genbank or fasta files of all the DNA sequences they have in their freezers/boxes, and this is indexed. You can upload a sequence that you have designed and want to build, and the tool will run designs against the index, developing assembly solutions that make use of the most efficient sources of template DNA across the network.

A list of fragment sequences from various "inventory sequences" is generated for any "design sequence", also identifying the smallest "unsourced" fragments you'd need have synthesised. You can request  sourced sequences from the people/groups who own them, and PCR primers are automatically designed to amplify the fragments from sourced sequences. Primers and fragments are designed based on your specific cloning chemistry choice. 

We have been showing the tool to academics and companies and getting a really good response. Its a pretty robust and pretty inventory management system on top of the solution generator, and I'd like to see if this is something the DIYbio community would like to start using. 

I think it would be really cool to have different hackspaces cataloguing their parts and making them all available to one another in this way. Maybe we could start getting some community projects going too.

Is anyone interested in making this happen?