Our fifth target is the E protein. We are emotionally attached to this protein because we explored its druggability in the early days of our project in the context of Covid.

At only 75 amino acids long, the envelope protein (E) is the smallest of the four structural proteins that make up the SARS-CoV-2 virus particle, and it is essential for the virus to infect cells. Data from other coronaviruses led researchers to suspect that groups of five E proteins form a pore that spans the lipid bilayer membrane of the virus. However, there was no direct evidence because the structural characterization of membrane-spanning proteins is difficult with the most commonly used techniques - X-ray crystallography and cryo-electron microscopy. Mei Hong's group at the Massachusetts Institute of Technology instead used nuclear magnetic resonance spectroscopy to solve the structure of the E protein and confirm pore formation (Nat. Struct. Mol. Biol. 2020, DOI: 10.1038/s41594-020-00536-8). The group also investigated how two drugs, amantadine, and hexamethylene amiloride, can bind to and block the pore. Although these drugs only bind weakly to the pore, the researchers say the new structural information could help in the development of drugs that target the virus.

In our HTVS, we will screen 1 billion of the small molecules to the pore of the E protein pentamer. We hope to find some nice molecules that can block the pore ;-)

---

How about start sending them out instead of waiting for every last straggler from T4 to return???

"Eight months into the COVID-19 pandemic, no vaccines or antiviral drugs are available against the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of the pandemic, owing to a lack of knowledge about the detailed structures and functions of the essential virus proteins."
That's a half truth at best. The structure of the Spike was known and vaccines were in trials.
Wondering what happens to a cell if your new E-blocking drug stops them from escaping the Golgi. Will the cell hopelessly clog up and die??? Will the cell be able to route all these budding failures to the lysosomes???

---

We switched to "corona_Eprot_v1" target. Tasks queue was empty about 5 or 10 minutes.

Sweet!!! I'm loaded up and they're running. Starting off estimating 1:16 minutes. :-)

---

As I can see, tasks for Eprot_v1 can be much longer. May be a good challenge for powerful crunch-chariots!

Hey hoarfrost,

I have some hosts that are taking about 2 hours to complete tasks, but other hosts with faster CPUs and larger L3 Cache per core, are scheduled to take over 5 hours. Is this normal and to be expected?

Is this what you mean by "As I can see, tasks for Eprot_v1 can be much longer..."?

Yes, for my Ryzen 3900X estimate time for corona_Eprot_v1 is about 5 hours.

Okay thanks

Dear Aurum,

Thanks for an interesting question. The Eprot is a multifaceted and still-enigmatic protein participating at multiple levels of the viral life-cycle. Regarding the Golgi, it is primarily located at the ERGIC when examining the cells but could be involved in multiple aspects of the viral replication cycle: from assembly and induction of membrane curvature to scission or budding and release to apoptosis (not hopelessly but gracefully die), inflammation or autophagy. It is definite that deletion of Eprot from SARS-CoV attenuates the virus whereas, in the case of MERS-CoV, virions are propagation deficient (virus can also partially compensate for Eprot deficiency but not very successfully). Eprot is mechanistically involved as a monomer, post-translationally modified, through heterologous and homologous interactions. Using our methodology we will try to study homologous pentamer and viroporin formation, if it can be de-/stabilised by small-molecules and its conduction properties modified. Is it related to pathogenicity? The research indicates yes, and through multiple pathways.

Just to mention one venue: Synthetic peptides corresponding to the full-length SARS-CoV E viroporin have been shown to be capable of transporting mainly monovalent cations as well as calcium connecting it to inflammation response. Inflammasome activation by CoV E was already reported and blocking ion channel activity with amantadine significantly inhibited activation of the inflammasome. This mechanism is also strongly connected to observed lung-damage. Interestingly upon mutation studies rendering viroporin ineffective, the virus very quickly regains pore activity mutations and restores its function. Thus, the research suggests that inhibition of the Eprot viroporin could limit pathogenicity and could be of therapeutic value. Similar studies were done on the M2 proton channel of influenza A virus with described rimantadine anti-viral properties.

All the best!

Marko Jukic
Natalia Nikitina

With the longer runtimes for the 5th target, lack of checkpointing is becoming more of a problem. I've had to remove these from machines that aren't always running 24 hours a day. I'll gladly put them back if checkpointing is implemented though...

Thanks, Chris

Thank you for the well-considered explanation. Virology is not my field, but it shows you have given significant thought to the problem.
That is a lot better than just choosing targets at random. I will try to keep a machine on it, more or less.
(I run my machines 24/7, and will try to make up for some of the people who have to have checkpoints. The longer ones are OK with me.)

Hi folks! We are extending the deadline up to 3 days. Thank you for CPU power!

Yes, for my Ryzen 3900X estimate time for corona_Eprot_v1 is about 5 hours.

But why are the faster cores taking longer? And to me it feels like its starting quite fast with 2% per Minute but slowing down to 0,42% per Minute and even less...
I'm not scared of longer runtimes, just the estimates are screwed up all the time...

Yes, for my Ryzen 3900X estimate time for corona_Eprot_v1 is about 5 hours.

But why are the faster cores taking longer? And to me it feels like its starting quite fast with 2% per Minute but slowing down to 0,42% per Minute and even less...
I'm not scared of longer runtimes, just the estimates are screwed up all the time...

It's estimation. BOINC client perform an estimation based on statistics of tasks computing on computer. And for new tasks it can be very inaccurate.

We generated a special, "experimental" batch of workunits with names started from "exp_eve". This bunch will be followed another one - "exp_rnd". Both batches use existing application and ligands library, but tasks from them scattered by search space.

With the longer runtimes for the 5th target, lack of checkpointing is becoming more of a problem. I've had to remove these from machines that aren't always running 24 hours a day. I'll gladly put them back if checkpointing is implemented though...
Thanks, Chris

I accidently tripped a circuit breaker and a rack of computers turned off. Lost several hundred CPU hours as they restarted from the beginning. We really need checkpointing for these big WUs. TIA

---