Hi all,
during these months of quarantine I read a lot about Rosetta@Home and the science grounding it. Of course my knowledge is very limited, but I am starting having a general intuitive understanding of what we are doing here.
I follow the names of the workunits sometimes, and I search on the internet for more informations about the proteins we are working on (when the WU name is explicative, which is not always the case).
Sometimes we work on proteins with no references at all on the Internet, such as JHR, b4k, b3x... Other times, I realized that we are working on well known proteins: bpmr (1,3), cd86, activin receptor, tgfr... If I look for a PDP structure of these proteins, I always found the crystal structure of them (and usually it was discovered several years ago). So I have some questions:
1) are we working on these proteins because the crystal structure is in some way incomplete and we need to improve it computationally?
1b) perhaps we need some kind of rosetta-model of these structures in order to work on them for example by designing proteins capable of binding to them?
2) are we doing benchmarks to see if and how fast the Rosetta algorithms can predict the actual structure?
3) are scientists manually tuning the algorithm in order to match with the already known structure?
4) is there already a kind of AI that analyze the results from Rosetta@Home and improves the algorithm so it needs to work also on already known crystal structures to train itself?

I am asking this just to understand how much are we working on "new" things, that is: de novo protein design, solving unknown crystal structures, protein-protein docking etc... and how much are we instead trying to improve the Rosetta algorithm and if scientists are doing it manually or by using an AI.

Reading a bit about the history of Rosetta@Home, I understood (perhaps wrongly) that during the very first years of the project the main goal were (2) and (3). Then, around 2017 if I don't misremember, David Baker announced a "breakthrough", that the Rosetta algorithm is now very precise and we can successfully design de-novo proteins.

Another small doubt: usually I cannot see the screensaver because it doesn't working on Linux. Now (and for a short period of time) I have a Windows 7 computer just beside me and sometimes I can see the screensaver. I noticed that for the cd86 workunits there was no "Native" box, while that box was present for the "JHR" workunits... and this was counterintuitive for me, because the cd86 native structure is solved while there are no reference at all on the Internet about JHR.

Thank you to answer my curiosity and sorry for my English :)

Very interesting questions.

P.S. I think recent wus that begins with b stands for "barrel"

---

Here are a couple of useful points that may give a partial answer:
https://boinc.bakerlab.org/rosetta/forum_thread.php?id=8128&postid=97365#97365
https://boinc.bakerlab.org/rosetta/forum_thread.php?id=14129&postid=98323#98323

---

‘JHR’ might be an abbreviation for ‘Junior HalfRoid’, which appeared in the name of a lot of work units earlier this year. There was a related Q&A here, though I’m not sure it answers your question…

While I have no knowledge of the ratios, you are definitely asking the right questions, and seeing the project in the right light. There are numerous groups in RosettaCommons that are working to "improve" the Rosetta algorithms. I put "improve" in quotes, because you have no way to know if your improvements work properly unless you test something using the original algorithm, and also using the new algorithm. If the result if that the new algorithm arrives at the same or more accurate answer using less models (less overall compute time), then it is indeed an "improvement", at least for the type of protein of your test.

The project is constantly asking the question "How could we have arrived at that answer, better?". Where "that answer" is perhaps from x-ray crystallography, or from running the latest R@h prediction. To answer that question, you constantly have to try new approaches against entire libraries of known structures. Your new approach might be better for some unique subset of the protein library you use for comparisons. Such as those that involve zinc, or have a hairpin curve, or are symmetric, or a barrel structure. Can we improve the predictions of barrel structures without harming our predictions of proteins that have hairpin curves? Until you study the predictions of your new algorithm against the various classes of proteins, you cannot fully understand what its predictions will look like. Then you ask things like, "can we retain the better hairpin curve predictions, without harming our zinc predictions?"

So, yes, R@h is really a project that is about making a better protein analysis tool, so that the tool is refined and working well when something like COVID-19 comes up. They also work with other researchers to test specific protein interactions, such as interactions with HIV, or malaria, or SARS-CoV-2.

---

Rosetta Moderator: Mod.Sense

So, yes, R@h is really a project that is about making a better protein analysis tool, so that the tool is refined and working well when something like COVID-19 comes up. They also work with other researchers to test specific protein interactions, such as interactions with HIV, or malaria, or SARS-CoV-2.

Something like this (new publication).

---

activin receptor, tgfr... If I look for a PDP structure of these proteins, I always found the crystal structure of them

If you press C several times in graphics window it becomes clear there are two proteins not one. Sctructures of tgfbR2 and activin_receptorII are already known so I suppose the second one is being designed de novo (probably for some anti-cancer stuff)

---