ELVES at Berkeley..rapid protein imaging

[sharp]

ELVES at Berkeley..rapid protein imaging

Ripped off the front page of the [H]

It seems that is is a "here is what these proteins look like" thing... in tune with Find-a-Drug in the "find stuff that works" compared to Stanfords "how and why does is work" approach that we need to understand this crap and design things to win permenately. ... VERY good stuff though, it makes solid data easier to get, and MUCH faster to boot. three cheers.

"Computers are faster today and people are automating many steps in the process of analyzing X-ray diffraction data, and Elves combines all these techniques together," Holton said. "Making the crystal is 90 percent of the effort, but once you have that and you get the X-ray diffraction data, you're 19 minutes away from solving it, whereas a couple years ago it took about a year."

and

Elves is based on a new kind of computer interface called the conversational user interface (CUI), invented by Holton. The interface interprets English-language commands, thus allowing the user to type in simple sentences such as, "Elves, I want you to process the data in my folder." It does this by using a search engine to search free-format text input by the user and associate numbers with program variables. The program restates what it will do, setting up a conversation with the user about the process.

it is not that long an article, and very cool. go read it

 

[Tigerbiten]

That the opposite of whats Stanford's doing.

There they are takeing an unknown protien.
Shineing the x-ray's through it and from the defraction pattern working out which atom goes where.
And henceforth working out the shape of the crystal/ protien.

In Stanford's case the atoms are known but the shape is not.
And as protien self organize, they tend to only give you one shape per sequence of atoms at a set temp.

Luck......