Finally, one thing that struck me is how friendly this conference and community are!

I sometimes write to the CCPEM mailing list and CryoSPARC forum, and generally try to answer questions when I can contribute something helpful. A few people told me they recognized my name from there and had appreciated something I posted. One person even told me they had been hoping we'd meet in person at the conference.
This was incredibly nice to hear!

I hope I'll be able to attend next year.

#CCPEM #CryoEM

One highlight from Friday that I forgot: Alister Burt's talk on his efforts to build #TeamTomo, a set of community-supported software packages and metadata standards for #cryoET. This is difficult work, and less rewarding for those in academia, but so essential. It's great to see progress on this front.

The other great news from Alister's talk: Warp is now supported on Linux!

Finally, I'll remember his advice "think about geometry, and use it if you can" when assigning initial particle orientations for subtomogram averaging.

#CCPEM #CryoEM

And some highlights from earlier today:

CryoCloud.io seems like a great single-particle analysis suite. They developed their own deep learning particle picker, which apparently outperforms Topaz. I hope this will be published, very curious to see how it works. It’s always great to have more options for particle picking (the most difficult and critical step of SPA in my opinion).
They also developed a metric to quantify resolution anisotropy without requiring particle poses (which are not part of the mandatory items for deposition, therefore rarely available).

Alexander Shtyrov from Garib Murshudov’s group showed very interesting results about learning atomic electron scattering factors from high-resolution cryoEM maps. This will hopefully help us use these maps to their full potential (pun intended) when doing atomic model building and refinement. The ideal training set for these methods are zero-dose extrapolated maps, which are still a rarity (there might be only one deposited to date? from Chris Russo’s HexAuFoil paper).

#CCPEM #CryoEM

These three days have passed by so fast, and here I am collecting my thoughts already on the way back home…

More highlights from Thursday:

Jessie Zhang showed great progress with the development of the Laser Phase Plate! Data collection with LPP on is now routine for both single-particle and tomography. It is still unclear whether fully in-focus imaging is possible; if so, it will of course change how we do CTF estimation, but the rest might work just the same? Unclear until they try it.

Jude Short in Chris Russo’s group has a working implementation of zero-dose extrapolation, presumably more user-friendly than the scripts that came with the HexAuFoil paper.

Hannah Bridges from StructuraBio (makers of CryoSPARC) reprocessed my dataset EMPIAR-10739 and got vastly improved results. This was so cool to see! Looking forward to seeing this case study in the CryoSPARC guide.

#CCPEM #cryoem

Couple of highlights from today Thursday (well, technically yesterday now, since it's getting late).

LocScale version 2 from Arjen Jakobi's group seems very helpful and I will try it as soon as I am back at the lab.

Doppio seems super nice, and has been made easier to install. I also need to try it. It would be nice to have an alternative to CryoSPARC, with a nicer workflow visualization than stand-alone RELION. Not sure what are the pros/cons between Doppio and Scipion, since they seem to do similar things. I need to read up about this.

RECOVAR from Amit Singer's group seems like a good tool to analyze heterogeneity. Main take-away from the talk: enough noise in the particle images can make most such programs "hallucinate" heterogeneity, so one must be very careful when interpreting results!

#CCPEM #CryoEM