For what it’s worth, I’ve spent the last few days going over most of the existing COVID-19 related structures and rebuilding/re-refining wherever I considered necessary. The resulting models along with some basic explanatory notes are at https://drive.google.com/drive/folders/1S5qJtCnK00NrcbwwBNgImUMewhiBkyPa?usp=sharing. Some commentary on things I found can also be found in my Twitter thread at https://twitter.com/crolltristan/status/1241317484235554823?s=21. Normally I shy away from meddling with other people’s new models without invitation - but if ever there was a time to make an exception, this is it. Please feel free to use these for whatever purpose you like, and to improve on them further if you can. Tomorrow I’ll start personally contacting the individual authors to see if they’re willing to update their PDB entries.
Wishing the best of good health to all! Tristan > On 20 Mar 2020, at 22:59, James Holton <jmhol...@lbl.gov> wrote: > > You might think that as a structural biologist you won't be able to do much > about COVID-19 anytime soon, but that is not true. Yes, real-world > therapeutics and vaccines take time, but we have already seen just how fast > we can get started. There are 21 PDBs already and some even have bound > ligands. Good job Frank et al. BTW! And my personal thanks to all of you > out there who are already hard at work on this. > > I believe this forum is an ideal place to share information and ideas on the > structural biology of SARS-CoV-2 as we move forward. It's a big virus, but > there are not that many proteins in it. If all of us independently do the > same bioinformatics and literature searches and end up trying exactly the > same thing in every lab all over the world, then that would be more than > unfortunate. To that end, I am personally interested on ORF8 for reasons I > will go into below. Has anyone tried to solve it yet? What happened? > Didn't express? Bad diffraction? What? Do tell. > > Some of us, as you may have heard, are stuck at home, our beamlines and labs > dark while we shelter-in-place. That doesn't mean our hands are tied. We > are still allowed to think. The fraction of the human race that has a > snowball's chance in Hades of figuring out this bug is very very small. > Structure may be your main skill set, but you are still a biologist. Do you > know how to run a PCR machine? Do you know how to pipette? You might think > that anybody can do it, but that is really not the case. Ever trained a new > student on sterile technique? How many days did that take? Now remember > that your student was no dummy and already studying biology. Everyone > reading this will make an excellent volenteer at the very least. I'm not > saying this to belittle the average human, only to say that we scientists, > moving in the circles we do, often forget that we have uncommon capabilities. > > For example, I also believe we can be useful in assay development. The void > left by the dearth and delay of test results has been filled with fear, and > that is a big problem. The tests, as defined, are straightforward, but also > extremely regimented like any good laboratory protocol should be. The US > CDC's instructions for academic labs are here: > https://www.cdc.gov/coronavirus/2019-nCoV/lab/index.html > My question is: how can this test be made faster, using more commonplace > supplies, in high-throughput mode and still valid? Not just for clinical but > for academic use? I think more than a few people on this list could be > regarded as experts in making a complex biochemical task faster, more > efficient, high-throughput and nonetheless valid. Yes, there are other > people who do virus testing for a living, but right now they are all rather > busy. Maybe if we put our minds to it we can help? > > As for why ORF8. I am basing my interest on the bioinformatics done in this > article: https://dx.doi.org/10.1093/nsr/nwaa036. Search for "T8517C" and you > will find what I'm talking about. The authors found two "types" of > SARS-CoV-2. They call them "S" and "L" because the only conserved amino acid > change involved is S84L in ORF8. The "S" type is believed to be the ancestor > of "L". What is interesting is how tightly linked this mutation is to a > silent mutation on the other end of the genome: the "L" type has a faster > codon for Ser in ORF1. Such tight coupling (r^2=0.945) means there must be > significant selective pressure preventing both of these mutations occurring > in the same virus at the same time. That, I believe, is interesting. > Espeically since they are so far apart I expect this selective pressure might > work in trans: as in a super-infection. That is, the S and L genome types may > interfere with each other. > > The authors fall short of claiming evidence of interference upon > super-infection, and indeed they have already been criticised for calling "L" > the "aggressive" type. But it is still interesting and points a finger at > ORF8. > > ORF8 has only one homolog in the PDB: 5o32 with 25% identity over a stretch > of 60 residues. This homologous region contains the S84L site (Val I544 in > 5o32). I had a quick look and appears to be a cavity-filling mutation to me. > Not very big, but maybe something could fit in there. To be sure we'd need > a structure of ORF8. > > Good luck to you all, and stay healthy. > > -James Holton > MAD Scientist > > ######################################################################## > > To unsubscribe from the CCP4BB list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 ######################################################################## To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
