Mike Robertson : Gemspot

In this episode, Ellen sits down with me to talk all things cryoDrgn ❄️🐉 Full transcript available at (https://github.com/bHimes/cryo2go/tree/main/transcripts/2020_EllenZhong_cryoDrgn)cryoDrgn is the software she developed during her PhD to model continuous heterogeneity in cryoEM data. She uses a combination of coordinate MLP and variational autoencoders to do so. Listen in if you'd like to know more about what that means! We also touch on some of the details that speed up the training of the neural network that never quite fit into normal academic talks, including the use of the Hartley transform. No cryo2go pod is complete without some time spent on career trajectories, and Ellen's is quite interesting... listen to the end to hear us riff a bit on academia vs. industry : )

In this episode, we get to hear from Dr. Christopher Barnes, currently at Caltech, about his research combining structural biology and immunology. Christopher was able to rapidly deploy his expertise developed in studying the HIV virus to the emerging SARS-COV-2 outbreak early in the pandemic, publishing this article in May 2020. See here in Cell (https://www.cell.com/cell/fulltext/S0092-8674(20)30757-1?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0092867420307571%3Fshowall%3Dtrue)This interview was recorded in June 2020 amid escalating calls for police reform and social justice in a country that has a long history of racial inequity. In the next episode, part II, we talk about how and when scientists should get involved in issues that many of us, myself included, shy away from.Between publishing 13 papers in 2020 and locking down an Assistant Professorship at Stanford starting summer 2021, I count myself luck that Christopher made time to sit down and talk with me. Enjoy the show!

In this episode, recorded a few days after the bioarxiv paper was released in March 2020, Dari expands on his work developing a set of biological priors using machine learning, which can then be used to regularize cryoEM reconstruction/refinement.We go beyond just the math and the methods and discuss the real what and why with respect to the motivation and prospects of this work."Exploiting prior knowledge about biological macromolecules in cryo-EM structure determination"https://www.biorxiv.org/content/10.1101/2020.03.25.007914v1orNow published in IUCrJ doi: 10.1107/S2052252520014384

We hear from Dr Kanika Khanna about optimizing cryo-FIB-ET and how she used the technique to advance our understanding of bacterial sporulation, as discussed in eLife:https://doi.org/10.7554/eLife.45257.001She also shared here outreach work with [Asha foundation | (https://sd.ashanet.org/)]Her website can be found here (https://khannakanika11.wixsite.com/mysite/)

Phase contrast provides the best means to high resolution electron microscopy images. The phase shifts induced by the light atoms that make up biological specimens produce very little contrast, which is why we usually image out of focus. This takes advantage of phase shifts introduced by the CTF at the expense of more complicated image processing, and loss of information near CTF zeros.Phase plates offer a potential solution, but have been hampered by difficulty in use (Zernike phase plate) and loss of signal due to additional inelastic scattering (Volta phase plate)Scwartz et al. demonstrate the use of very intense (the most intense!) laser light to induce a selective phase shift enhancing image contrast in moderate energy (80 KeV) electrons.The project web page can be found at:http://matterwave.physics.berkeley.edu/phase-contrast-electron-microscopyThe paper discussed (Laser phase plate for electron microscopy) can be found: https://www.nature.com/articles/s41592-019-0552-2A follow-up paper detailing the physics behind the laser:https://arxiv.org/abs/1910.14441

The first full episode. Mike Roberston tells about GEMSPOT, a pipeline for modelling ligands into their binding pockets that integrates constraints from the cryoEM map with traditional computational chemistry methods. The pre-print can be found on bioarxiv at doi: http://dx.doi.org/10.1101/750778Mike is on twitter @BiophysicsGuy

I recorded my first interview a couple of days ago and have another scheduled for Saturday. I'm excited to get this thing rolling, so here's the intro, fresh off the mixer