Thanks for your comment and perspective. I wasn't aware JEOL and others had 100 kV competitors suitable for imaging biological samples under cryo conditions.
The situation with Thermo Scientific is certainly concerning, and one could imagine a scenario similar to where the large software companies reportedly hire workers in part just to starve the competition of talent. (In Europe, lack of engineer availability is a major complaint I've heard from CRYO ARM users, which could fit within this framework.) Thankfully the UK blocked Thermo's attempted acquisition of Gatan, and now with the Apollo there will at least still be competition in the market for detectors.
It sounds like the situation with EPU/SerialEM has some similarities to the situation in data processing for single-particle reconstructions with cryoSPARC vs the open academic software packages. Not just the use of open licenses for code, but decades of algorithm development and open discussion in the literature clash with the potential to use patents, closed formats, and general lack of interoperability to give a single provider an effective monopoly. Maybe simply having commercial funding for UI development and testing will be enough to drive most (new) users to the commercial platform. One positive development in this area was the recent release of software for motion correction and CTF estimation under permissive licenses (https://github.com/czimaginginstitute). However the trend, at least in PDB depositions, is decidedly in favour of closed-source commercial software.
> Thanks for your comment and perspective. I wasn't aware JEOL and others had 100 kV competitors suitable for imaging biological samples under cryo conditions.
Just to be clear, I think the Tundra's market ("entry level" or about 1.5 million) is actually relatively under-served. The bad behavior is Thermo coupling their entry level microscope with a bunch of software restrictions that will prevent adversarial interoperability with other vendors, something that has not been the case in the past. Their previous entry-level line, the Tecnai Spirit 120 kV, had a full scripting package and could run SerialEM/Leginon, etc, and use Gatan or TVIPS or AMT or any other camera package.
>It sounds like the situation with EPU/SerialEM has some similarities to the situation in data processing for single-particle reconstructions with cryoSPARC vs the open academic software packages. Not just the use of open licenses for code, but decades of algorithm development and open discussion in the literature clash with the potential to use patents, closed formats, and general lack of interoperability to give a single provider an effective monopoly. Maybe simply having commercial funding for UI development and testing will be enough to drive most (new) users to the commercial platform. One positive development in this area was the recent release of software for motion correction and CTF estimation under permissive licenses (https://github.com/czimaginginstitute). However the trend, at least in PDB depositions, is decidedly in favour of closed-source commercial software.
It was nice to see MotionCor and AreTomo go open source. Generally, I'm less concerned about the reconstruction side because the file formats are pretty well standardized (despite being awful), so there's no lockout. I love RELION and EMAN2 dearly, but, particularly RELION needs to throw a FTE at a UX designer. CryoSPARC is just so much easier to use, easier to manage, easier to onboard folks. RELION really is a nightmare of complexity for new users.
You seem like you have quite a bit of experience - I'm curious what your background is?
> I love RELION and EMAN2 dearly, but, particularly RELION needs to throw a FTE at a UX designer. CryoSPARC is just so much easier to use, easier to manage, easier to onboard folks.
I've heard that the CCP-EM is working on a new front-end for RELION. It's doubtful that it would reach the same ease-of-use as cryoSPARC, but it might be a step in the right direction.
> You seem like you have quite a bit of experience - I'm curious what your background is?
I'm a freshly minted postdoc. I studied biochemistry and somehow spent about half of my PhD processing single-particle cryoEM data.
The situation with Thermo Scientific is certainly concerning, and one could imagine a scenario similar to where the large software companies reportedly hire workers in part just to starve the competition of talent. (In Europe, lack of engineer availability is a major complaint I've heard from CRYO ARM users, which could fit within this framework.) Thankfully the UK blocked Thermo's attempted acquisition of Gatan, and now with the Apollo there will at least still be competition in the market for detectors.
It sounds like the situation with EPU/SerialEM has some similarities to the situation in data processing for single-particle reconstructions with cryoSPARC vs the open academic software packages. Not just the use of open licenses for code, but decades of algorithm development and open discussion in the literature clash with the potential to use patents, closed formats, and general lack of interoperability to give a single provider an effective monopoly. Maybe simply having commercial funding for UI development and testing will be enough to drive most (new) users to the commercial platform. One positive development in this area was the recent release of software for motion correction and CTF estimation under permissive licenses (https://github.com/czimaginginstitute). However the trend, at least in PDB depositions, is decidedly in favour of closed-source commercial software.