Just finished our monthly call with the research team.

1. We discussed the most recent rate of batches per day and the backlog (see below for the latest stats). The research team is building more batches, which we should have by the end of this week or beginning of next week. They're happy with the current pace.
2. We pointed out a recent question elsewhere in the forum about MIP. One of the researchers is going to answer the question himself.
3. They're continuing work on the papers we've mentioned in previous updates. One paper was not accepted by the first journal they submitted to (although it got good reviews, the human microbiome is not top-of-mind for many journals right now), so they're prepping it for another journal. An additional paper is almost ready for submission.

Current status of work units:

Available for Download: 775 batches
In Progress: 9,007 batches (16,199,030 work units)
Completed: 301,472 batches total - 5,604 batches in the last 30 days - average of 186 batches per day
Estimated backlog: 4 days

Thanks for the update.

I think a good question that should be asked the researchers is to tell us more about the 1/3 of WUs that were "unsatisfactory": if they figured out why, if they solved the problem and what they think about the fact that they are using a very old version of Rosetta.
Regarding this last point, I am not in the position to criticize any software choice of the research team: I just wonder if there is a reason or if it is just because it works well and they are happy with it (thus, the old software is not the cause of the lost wus).
I really like this project, but I am not very happy to crunch it if I know that one third of my electricity and computing time goes into nothingness.

Also, It would be nice if we could know something about the progress of this project. We know that they solved more than 300.000 protein structure, and this is very nice, but they speaks about "trillions" of bacteria. This mean that if we are going to solve all of them this project is just infinite... Also because by reading the news I got to know that this is just one step in a very ambitious project that requires:
1) Machine learning with results coming from us and crystal structure (already in progress on their machines);
2) Protein docking to see how they interact (... perhaps a "phase 2" here?)
3) More and more research

It would be nice to know at least how far are we to solve all the protein structures they want to be solved.

Thank you :)

Ok, this clears a lot of things, I am sorry for my misunderstanding.
Yes, I am aware that unexpected results are fundamental in science; I am aware also that even "bad" results can be actually "good" for a lot of things. I was mislead by reading how other users reacted to your May update: it looked like they just realized after three years 25% of the WUs were to be discarded but indeed by rereading your post it is as you said now.

I wish the best for points 1, 2, 3 and 4.
I keep the other two questions open:
1) why are they using an old version of Rosetta? Not to criticize, just for curiosity if there is a scientific explanation (perhaps their ML software is trained with that software, or they made some modifications to it)
2) Is it possible to assess, even in terms of order of magnitude, the amount of work we still need to do to complete this phase of the project, that is, protein structure prediction of humane microbiome? I mean, do they add proteins as long as the research goes (and so we can not know in advance how many of them we need to solve) or do they already know the number of protein? I remember some years ago the very nice "percentage of completion" info of the project. Would it be possible to do the same for this one?

thanks.