3 ms·

Hey amelius! Great questions. The target audience for these instruments are scientists and hobbyists who don't want to spend a ton of money on similar commerci

by sinab 8y ago

Hey amelius! Great questions. The target audience for these instruments are scientists and hobbyists who don't want to spend a ton of money on similar commercial systems, and want the flexibility to modify their operation (e.g., poseidon can run custom flow profiles per experiment whereas off-the-shelf commercial systems typically only run one flow rate per experiment). The purpose of poseidon is to show that open source biological instruments can be developed and used by a community, similarly to how open source software tools in biology are developed and used. Academic budgets vary from institution to institution and are sometimes determined by exogenous forces beyond the lab's control. A complete commercial system to do single-cell RNA sequencing costs tens of thousands of dollars! Using alternatives such as the Harvard Apparatus syringe pumps and DropSeq [0] to run the same experiment will still cost you into the thousands of dollars. With the poseidon system, we greatly reduce these costs. Users can build the instruments to run these experiments for less than $400 and are not restricted to additional costs and tedious firmware upgrades to expand the system. In response to your point on time management and instrument-making, I think that if there exists a need to develop these systems such that they will advance biological experiments then it's totally cool to have academics work on these sorts of projects! Biologists and bioengineers have always developed tools alongside discovery and this is no different from developing bioinformatics tools. [0] https://www.cell.com/abstract/S0092-8674(15)00549-8 https://www.cell.com/abstract/S0092-8674(15)00549-8

amelius 8y ago
Thanks for these great answers. The only point where I would disagree is your comparison with bioinformatics software. The difference there is that installing a bioinformatics package and learning to use it takes far less time than crafting an instrument, especially considering tasks like finding the right suppliers of components, and calibrating and testing the device (for which you might need other instruments). But on the other hand, perhaps building these self-made instruments can become a task for specialized, facilitating instrument-makers that are already part of many institutions.
- munfred 8y ago
  Note: I also helped develop poseidon The poseidon system was explicitly designed with ease of assembly in mind. If you look at the build videos [1] you'll see that assembly of the entire system (3 pumps + microscope station) takes less than an hour and requires just pliers and screwdrivers. The importance of ease of assembly was a lesson that we took from assembling the miniDrops microfluidics station [2] developed specifically for one kind of experiment (dropSeq) [3]. The miniDrops is very good at what it does but assembly was somewhat cumbersome: it required ordering a custom PCB, specialized parts only available from one vendor (whom I had to nag over the phone to send me a quote!) and assembly of the device itself took 10-20h. Not every kind of equipment can be made as easy to source and assemble as we did with poseidon, but we really think that keeping this at the front of your mind can make or break the adoption of a piece of open source hardware. This is especially true in the context of biology laboratories, where many people are not what you could call "hackers" or "makers" and will be immediately put off by a daunting assembly process. [1] https://pachterlab.github.io/poseidon/hardware https://pachterlab.github.io/poseidon/hardware [2] https://metafluidics.org/devices/minidrops/ https://metafluidics.org/devices/minidrops/ [3] https://www.nature.com/articles/s41467-017-0265 https://www.nature.com/articles/s41467-017-0265

by sinab 8y ago

amelius 8y ago

Thanks for these great answers. The only point where I would disagree is your comparison with bioinformatics software. The difference there is that installing a bioinformatics package and learning to use it takes far less time than crafting an instrument, especially considering tasks like finding the right suppliers of components, and calibrating and testing the device (for which you might need other instruments). But on the other hand, perhaps building these self-made instruments can become a task for specialized, facilitating instrument-makers that are already part of many institutions.

munfred 8y ago
Note: I also helped develop poseidon The poseidon system was explicitly designed with ease of assembly in mind. If you look at the build videos [1] you'll see that assembly of the entire system (3 pumps + microscope station) takes less than an hour and requires just pliers and screwdrivers. The importance of ease of assembly was a lesson that we took from assembling the miniDrops microfluidics station [2] developed specifically for one kind of experiment (dropSeq) [3]. The miniDrops is very good at what it does but assembly was somewhat cumbersome: it required ordering a custom PCB, specialized parts only available from one vendor (whom I had to nag over the phone to send me a quote!) and assembly of the device itself took 10-20h. Not every kind of equipment can be made as easy to source and assemble as we did with poseidon, but we really think that keeping this at the front of your mind can make or break the adoption of a piece of open source hardware. This is especially true in the context of biology laboratories, where many people are not what you could call "hackers" or "makers" and will be immediately put off by a daunting assembly process. [1] https://pachterlab.github.io/poseidon/hardware https://pachterlab.github.io/poseidon/hardware [2] https://metafluidics.org/devices/minidrops/ https://metafluidics.org/devices/minidrops/ [3] https://www.nature.com/articles/s41467-017-0265 https://www.nature.com/articles/s41467-017-0265