At present I have 37 ab tasks in line. In ab_07_20_T0618_d003_68_172_d003__SAVE_ALL_OUT.IGNORE_THE_REST_09_15_21680_260_0 can someone explain what the 07_20 & 09_15 & T0618 means?

Another example of numbers is ab_07_19_T0616_d000_h002__SAVE_ALL_OUT.IGNORE_THE_REST_04_05_21665_130_1 07_19 & T0616. Would I be correct in saying 07_20 is a date?

Can a Mod please rename this thread to How Are Task/Work units Named? Thanks

---

Have a crunching good day!!

Thanks transient. I wonder how many target there are for CASP9?

---

Have a crunching good day!!

Here is the list of CASP9 targets released so far. According to the CASP9 thread on this forum we will probably be crunching until the end of August (4 months from the start of May), so there probably aren't too many more to go.

You may notice that 5 targets were cancelled as the real structures were released during the experiment. Early release is usually due to a separate group of researchers working on the same protein that the CASP researchers worked out.

---

Regarding CASP9, is this like a "competition" against other protein predicting "software" out there?

---

CASP basically helps discover what the current state-of-the-art is in the field. After the exercise, the labs that produce the best results are invited to present at a conference and share information about their techniques to help advance the state-of-the-art.

Some of the predictions are done in software, as with Rosetta@home, but there is another class of predictions that are done by people (with the aid of software, but keeping the human intuition in control).

---

Rosetta Moderator: Mod.Sense

Regarding CASP9, is this like a "competition" against other protein predicting "software" out there?

CASP is held every 2 years with CASP9 being the 9th. Officially it is an experiment to advance the science of protein prediction by seeing which methods work well and sharing knowledge. Unofficially it serves as a competition to see which research groups are doing the best and benchmark the different systems/methods.

Some research groups use purely human prediction methods, while others rely to a greater or lesser extent on the support of computer systems. For this CASP there are "server prediction" deadlines for using pure computer calculations and "human prediction" deadlines for where humans are involved in the prediction process.

The Bakerlab team is submitting answers in both categories. The server deadlines are tight so the tasks are distributed through Rosetta@home and the best answer that comes back within 18 hours is submitted to CASP. The human deadlines are a few weeks longer so tasks are sent out through Rosetta@home then analysed by the scientists, which allows them to judge if a result looks good or bad.

It will be interesting to see not only how Rosetta compares to other systems and methods, but how the purely automated "Server" results for Rosetta compare to the "Human" results for Rosetta. The ultimate goal would be to see the quick, automated results beat or equal the slow, human ones but that is perhaps not a realistic goal for this season of CASP.

---

Interesting information thanks. In ab task names there is IGNORE_THE_REST. Why is this in task names? Should we manually press update on our clients to return completed CASP9 tasks or let boinc do it's own thing?

In ab task names there is IGNORE_THE_REST. Why is this in task names?

Don't take this literally. What it probably means is that they have a protein where they want to see what happens if they alter only one section. Therefore they make the rest of the protein "inactive". By testing different methods they can try to devise algorithms that both produce good results and reduce processing time.

---