Hi Rosetta@home community,

I've been recently submitting jobs of what appears to be a large membrane protein. These have many helices running up and down vertically. I hope these are not slowing down your computer... too much

The membrane protein, I'm working on has no structure or homologous structure, so this is an "Ab initio" job.

We have an idea of which parts of the gene are transmembrane and which parts are outside the membrane, but how the helices are arranged is unknown.

We also have some information from evolutionary constraints, derived from the study of how his particular family of proteins evolved. These constraints can be used in Rosetta to reduce the search space. But we still need your computing power to sample the smaller space!

If you have any questions, feel free to ask. I'll keep you updated as to what we find.

Thank you for your time (and please thank your computer for me)

-krypton

Thanks for the post: always interesting to know what we're working on. What are the names given to these jobs?

Thanks for the post: always interesting to know what we're working on. What are the names given to these jobs?

Hi svincent,
These might not have the most friendly names... But they start with "zinc".

Ah, I see a couple of those in my queue. I concur, always nice to know what we are working on.

Thank you Kenneth.

The exciting part is that we are actually converging on an answer! I sent the models to an expert that works on this protein experimentally to see what he thinks.

Little more about the protein in question: It is a human zinc transporter protein. A mutation in this protein is known to cause a genetically inherited disease: Acrodermatitis enteropathica. It has also been shown to be involved in human pancreatic cancer

Membrane proteins are very difficult to solve experimentally... some are impossible to solve without heavily modifying the original sequence or adding a stabilizing domain of sorts. If we can computationally solve this, it should give a boost to folks working on the gene.

I increased the number of jobs to see if we get better models.

-krypton

Hi all cruncher,

I'm excited to announce that the structure you guys predicted for hzip4 is now published and is being used by scientists to design new experiments! We used nearly 100,000 cpu hours for this project.

Check out the following link:
http://www.jbc.org/content/early/2015/05/13/jbc.M114.617613.short
I'll update the link once the final paper is out (the link above is for an early edition).



To download the final model see here:
http://gremlin.bakerlab.org/hzip4.php

Thanks for the info krypton! It's this feedback that keeps people running/installing rosetta :D

---

Congrats on the new publication Sergey! ... and the needle of science moves forward another tick ;-)

Fantastic!

---