ReSeq: Repurposing HiSeq DNA Sequencers | Hacker News Reader

ReSeq: Repurposing HiSeq DNA Sequencers

reseq.hackteria.org

59 pointskasbah21 comments

Mikeselectricstuff has some teardown videos of DNA sequencers. Interesting stuff.

https://www.youtube.com/watch?v=XaumUp4GpCw

I am part of this project but just about to board a 2 hour flight. If you have any questions, feel free to post below and I will answer them later.

This is very cool. I hope you get some cool uses out of this.

What's the max frame rate you can get out of the cameras (without tiling)? Are you able to control focus and camera gain?

Thanks, working out the max frame rate is a bit complicated by the fact it uses time delay integration (TDI) scanning. There are 4 TDI cameras each 2048 * 128 pixels. I couldn't find any official documentation (the cameras are Hamamatsu S10405 CCDs and there is one C10000-509 control board for 2 cameras each). I think it's the same sensor as this camera which quotes the line rate at 50khz. https://www.hamamatsu.com/resources/pdf/sys/SCAS0085E_C10000...

This Ebay listing quotes the line rate as 1khz, not sure where they got that: https://www.ebay.com/itm/4-Illumina-HiSeq-2000-Hamamatsu-S10...

We were able to grab a live video feed from a single camera, that must have been 20fps at least, but we'll have to investigate fully.

Yes, we are able adjust the gain and are able to focus both by moving the stage and by adjusting the more fine-grained piezo focus actuator (but we'd like to be able auto-focus using Micro Manager: https://micro-manager.org).

Nice. I suspect there are quite a few Illumina engineers secretly cheering for you.

I'm in touch with a community of microscopy folks which includes the maintainers of Micro Manager. I'll send this link around, I bet they'll be interested.

I saw an ebay link for 4 cameras pulled from a HiSeq 2000. I am interested in using the cameras to build a digital microscope, but I'm not sure how easy it would be to interface the cameras to a PC, or if they need optics to work. Any ideas? Here is the link: https://www.ebay.com/itm/4-Illumina-HiSeq-2000-Hamamatsu-S10...

What about reagents? The sequencers themselves are expensive, but library kits cost a lot of money, as do the reagents for running a flow cell. Even if I can get a sequencer for free, if I still have to pay through the nose for reagents it's no good. Especially as these sequencers become obsolete and people stop MAKING the relevant reagents, and their cost goes higher and higher.

You can make your own reagents if you want to do sequencing reactions on this. It's not inconceivable.

It's not necessary even to do sequencing. These systems are at their heart a general purpose microfluidic system, temperature controlled flowcell with high throughout fluorescence imaging capabilities. You can do a hell of a lot of stuff with that feature set.

Good question! I've always wondered what would be required (in terms of equipment and raw materials) to make those reagents, I'm really curious to know how those get synthesized to and the level of purity/quality required for a successful sequencing run.

I think the idea is to use these as "automatic microscopes" instead of as sequencers.

I'm also involved in the project. Ideally we'd like it to be able to be used both as both a microscope and as a sequencer. Potentially there are other uses too. For example this paper uses a customised Illumina GAIIx sequencer to do high throughput measurement of on and off rates: https://www.pnas.org/content/114/21/5461 While this does involve sequencing with Illumina reagents, my point is that it's possible to use customised flow cells and reagents for new assays that you might dream up. Software to run such assays could be built on top of what we are doing.

I have heard that the BGISEQ-500 (https://www.bgi.com/resources/sequencing-platforms) is a similar machine...

Our initial focus is on writing the device adapters that enable Micro-Manager (https://micro-manager.org) to talk to the HiSeq hardware. MM will then be able to drive it like a microscope as well as providing access to all the other fluidics switches, pumps, sensors etc. The MM user interface is extensible and plugins can be written to support assays such as sequencing. MM also exposes all of the devices and functionality in a scripting environment.

Here's another series of posts where they tear down one of the original Illumina sequencers (the Genome Analyzer).

https://blogs.swarthmore.edu/Illumina+GAIIx+Teardown/

Ooh. I got a HiSeq 2000 for a six pack of beer. Can’t wait to try to hack it!

You should become an alpha tester!

I just nabbed a cheap one as well. Happy to join the community.

Question for you though: it appears to be just an inch too big to fit into my building. How difficult is it to split into manageable chunks. Do the 4 modules separate without too much hassle, or are we talking full teardown?

It's pretty modular. Urs Gaudenz removed the fluidic pump and fridge on his to reduce the size.

sweet! I was hoping it could be disconnected down the middle. I'll have to reach out and see how involved that was.

Where's the best place to chat with the community? Got a slack channel or anything?

In the forum currently: https://forum.hackteria.org/t/hiseq2000-next-level-hacking/3...

You can cut it down the middle but there is quite a lot to move out of the way first. The microscope and power supplies are on the left and the rest is on the right. If you are only an inch out then it may be enough just to take the plastic panels off.

Boy, do I feel old. HiSeq 2500 was a gold-standard, $750k machine when I was starting grad school. Which was not that long ago, or so I thought.

Reply on news.ycombinator.com