Video Tip of the Week: Aquaria, streamlined access to protein structures for biologists

This week’s Video Tip of the Week is Aquaria, a new resource for exploring protein structures, mutations, and similarities to other proteins. It’s a very well-designed and interactive experience for end users. It is aimed largely at biologists who could benefit from exploring the structural details of their proteins of interest, but are daunted by tools aimed at structural biologists. But for tool developers, you should also look at how this rollout went. It’s one of the best examples of a tool launch I’ve seen in this field. And I’ve seen a lot.

So first, the tool. Aquaria offers users a streamlined way to access and explore protein structures. Combining the kinds of information you get from the PDB structure resources, and additional details like the UniProt mutations. Currently you start with a basic search by asking for a protein by name, or PDB or UniProt ID. They have pre-calculated the relationships of proteins in PDB and Swiss-Prot to quickly offer you a structure and related proteins. The paper notes: “Currently, Aquaria contains 46 million precalculated sequence-to-structure alignments, resulting in at least one matching structure for 87% of Swiss-Prot proteins and a median of 35 structures per protein….” In addition, it lets you explore other important biological features such as InterPro domains, post-translational modifications, so you can think about how the mutations + structures + functions impact a given protein that you are interested in. As they describe it:

“We have loaded SNP data from Uniprot and Interpro so you can see where the mutations lie on your 3D model. And we have found that you may be pleasantly surprised to find your mutations clustering in 3D space!”

The Aquaria folks provided an intro video to get you started:

Another handy feature they provided is a Quick Reference Card with shortcuts to the functions [PDF]. In addition to this intro, they have a longer video as well. This is more like a typical lecture with the background, the framework, the goals of the project, and more about the underlying database.

Now, this thing about the rollout of this software project. I found it when I was looking over the talks at the upcoming VIZBI conference (Visualizing Biological Data). Every year I find there are awesome ideas that come out of VIZBI, and tools I want to explore. Among them this year is Aquaria. So I went looking for more detail, and found some of the traditional stuff. The paper (below), the press release, etc. And then I found the Reddit discussion. The Aquaria team did a Science AMA on this tool. It engaged a range of folks–some folks just fans of science who had probably never seen protein structures before. That’s fine with me–the more folks who appreciate research and learn about how researchers explore proteins is a good thing. But others had good technical questions for the team–such as other ways to find proteins of interest with sequence searches, or integration with other tools like UCSC Genome Browser. All the answers are over there. I enjoyed the question about the name of the tool:

It seems you get the ideas we had in mind: using Aquaria lets us observe these fascinating creatures (proteins) from the natural world. Aquaria creates an artificial environment and lighting where we can observe isolated proteins; like aquarium fish, proteins are often beautiful and (usually) live in water.

I asked them about how this played out, and they had ~1000 folks visit their site as a result of this Reddit event. That was really interesting to me, and a very neat route to drive awareness.

They also provided a way to support users with one of my other favorite resources–Biostars. They created a support thread there where uses can ask questions and get answers. I so prefer this to mailing lists, and I’m glad to see this easy method to get support. In fact, I asked something that I couldn’t quite figure out yet.prot_structure_sample (Here’s the protein I was looking at: I wanted to see all the subunits in full color, you de-select autofocus to do that. And color by chains for this version.)

Also, for the developer types: they offer a way for you to interact with the Aquaria software to add your own features of interest with their API. Maybe you have new mutations you have found in some sequence you’ve obtained in your lab, for example. They are offering guidance on that here: They touch on this in the longer video (~27min) if you want a bit more explanation. I suspect from the high quality support they are offering, they’d be interested to hear from you and what features you’d like to see applied to these proteins as well.

So kudos to this team for a nifty tool and really serious multi-media outreach efforts. I think it was well done on all counts. I’ll bet you Reddit reached more of the right folks than a press release ever will. PIOs take note–get your scientists on Reddit.

Quick links:

Aquaria site:

Reddit Science AMA:

Biostar support thread:

O’Donoghue S.I., Kenneth S Sabir, Maria Kalemanov, Christian Stolte, Benjamin Wellmann, Vivian Ho, Manfred Roos, Nelson Perdigão, Fabian A Buske, Julian Heinrich & Burkhard Rost & (2015). Aquaria: simplifying discovery and insight from protein structures, Nature Methods, 12 (2) 98-99. DOI: