Folding Forum

I'm new to this amazing project.
Do the FAH servers, which receive a finished WU, check if its calculations are correct?
In other words, is "folding" a protein an NP-complete problem?
How could you check the WU?
Thank you very much!

Welcome to the forum TheeeeGod

Yes, the FAH server perform a check on returned WU to ensure the result is consistent and scientifically correct. Sometime they send you back that a result is not liked by the servers. Also during the folding on your hardware the system is performing checks and cause sometime that a WU get dumped.

In all cases you can post here the log of effected WU and a moderator can dig deeper and find out if that problem happen also with others donors assignments. They can additionally contact the researcher in case of further investigation.

This way of validation is a big difference to BOINC where often results are verified in peer-assignments, causing redundant calculation of the same WU. Some people (incl myself) think that's waste of resources.

Good question. Short answer: No, folding isn’t NP complete (otherwise biology couldn’t exist). Checking WUs is a longer and pretty technical discussion.

Thank you very much for your responses.
I just want to make sure I understood:
What is the ratio of the computational resources required to finish an WU to the computational resources required to check it?
In other words, relatively to folding, how hard is the checking?
Thank you again!