synthetic protein design in virtual reality

Case number:699969-2003201
Topic:General
Opened by:scottyler89
Status:Open
Type:Suggestion
Opened on:Saturday, December 10, 2016 - 18:25
Last modified:Sunday, February 26, 2017 - 16:03

This is probably getting a bit away from the foldit specific aims - but a tool that I would really love to have as a biologist, is a virtual reality implementation of foldit for synthetic biology. For making protein fusions, and things like that - it could really help synthetic protein design to have a rapid in silico prototyping tool like foldit in VR. Similar to this vein is VR for pymol (https://www.virtalis.com/vr-for-pymol/), but it doesn't have the modeling tools to compete with Foldit. Thanks for the consideration!

(Sat, 12/10/2016 - 18:25  |  3 comments)


Joined: 01/09/2017
Groups: None

Are there any plans to port foldit to VR (HTC Vive in particular)?

Joined: 04/24/2014
Groups: None

As always, our top priorities in development are aimed toward improving Foldit for hardware currently available to users. While we are a small development team, we are always open to the possibility and potential for graduate students and funding towards increasing development energy (and staff) on and for new platforms and input devices.

We appreciate the suggestions and discussions that may assist (but not guarantee) prioritizing the types of hardware support that would be most useful to the community!

Joined: 09/29/2016
Groups: Gargleblasters

I saw that Dun-Yu Hsiao (a FoldIt Dev) is working on a Microsoft Kinect version of FoldIt [The 'Kinect' is an xBox game console device that ultimately was released for PC, due to the "hardware hacking" crowd making genius use of it for many other purposes on the computer. Its role in either case is the detection and interpretation of user's physical movement to convert into gameplay or interface manipulation. In short, it makes your hands, arms, eyes, etc capable of doing things typically assigned to your mouse or even keyboard]. I would argue that the Kinect is just as available to users (ie: Foldit Players) as VR, but in reality (no pun intended), VR might actually be even more available.

I base this on multiple factors (listed in order of cheapest to most expensive):
- First, the "Google Cardboard" API, which takes any of a wide number of supported Android phones (and later on I think even iPhones), places it in a head mounted structure/case that houses two lenses, thus creating a VR headset. Named as such due to originally being made simply out of cut and/or folded pieces of cardboard. While this is old and has fallen out of favor for newer things, it is till arguably the cheapest option, ranging from a few dollars for the Cardboard kit, to even only $10 for a molded plastic version (which a friend has and said works rather well.)

- Second and Third I'll combine since they are, again, Android-phone based solutions --which frankly, proves my point further that it's widely available due to pretty much everyone owning a smartphone these days-- called Google Daydream (for their Pixel smartphones) and Samsung's GearVR (for their Galaxy S and Note devices). Daydream basically is the successor to Cardboard, but as far as I know, only supporting the Pixel handsets. GearVR, while again locked to certain devices (Samsung in this case), has the added benefit of being backed by the now-Facebook owned company Oculus [sic], who make the more widely known Oculus Rift. Which means the GearVR either has support for most of, or all of, the content available created for the actual Oculus Rift VR headset; a sizable compendium of applications of various types. These phone based devices are very similar to the molded plastic Cardboard ones mentioned above, but with far more quality and comfort. Their prices range from $65 to $130, which is still quite affordable.

- Fourth, the actual, and numerous, dedicated Head Mounted Displays (HMDs) which these days are generally referred to "VR Headsets", but also include devices that are not VR and simply are a small screen fitted to your face (due to how content is displayed, it mimics the same feeling as if you were viewing it on a movie theater screen). This include the two more well known names, the Oculus Rift and the HTC Vive (also sometimes may be called the Steam Vive, as it was commissioned by the game distributor Steam www.steampowered.com). However, there are many many other models on the market, some with far better features and specs, some with not, others that are simply on par. Arguably, the Vive is the leading product with consumers at the moment, due to the inclusion of multiple motion sensors and handheld controllers which do the same things that the aforementioned Kinect does. Worth noting is that the Rift provides similar functionality, simply in a less effective manner in some instances. Prices on these however, are many times more expensive than the previous options, in multiple regards. Namely their purchase cost which is around $500-$650, but also due to the fact that these are simply screens on your head with sensors and that means you still require a computer that is capable of rendering the content.

Thankfully Foldit is not very graphically intensive, even if the current state of coding is less than optimized for most, which results in lackluster performance with the game opened/visible (versus minimized). By comparison, the smartphone based options require slightly less, depending on how they're being utilized. If the program is fully ported to Android (something I thought would've long been done by now), then it would be running on the phones, and I think they'd handle them fine. People would need to have a Bluetooth keyboard AND mouse to connect to their phone in order to interact effectively and efficiently, but those are inexpensive items. If they are used as the "screen" and connected to the computer, then one faces the same situation as using the HMDs like the Vive or Rift, a task I figure most $500+ laptops should be capable of handling provided Foldit receives the much needed attention to its coding (lord knows there are plenty of users in the community here who are not only more than capable of lending a hand, but more than willing to).

Combine all of those things and we have quite a list of possible outputs for this VR Foldit to reach, and with the Playstation 4 now also having a VR headset (and comparatively affordable vs Rift/Vive), you would tap even more people. Which given the PS4's hardware specs, it would more than easily handle Foldit (7 usable CPU cores, ~7GB of usable RAM, and a respectably powerful graphics chip).

Best part of all? With Kinect support already having been worked on, I think the step to VR is not only the most logical choice, but likely not as difficult had all that work not already been done. I suspect that the majority of time would have been spent coming up with ways to interact with the game interface without any anything but hand movements and gestures.

Even if that element is not utilized, mostly all that needs to happen to get it working for VR headsets is a render path that outputs what is being display, but just slightly offset L and R. This is generally why the VR games require a higher end computer than the game would typically need, because everything is in effect being rendered twice; however, in Foldit's case, there isn't much that gets rendered and so the additional workload would be minimal. I figure, not being a coder myself, the easiest approach --albeit, definitely not the most computationally efficient-- would be to basically run a second instance of Foldit on another thread (given the game isn't very multithreaded to begin with) which everything is being mirrored, except the GUI for each is forced to being shifted and perhaps rotated a minuscule amount to compensate for eye spacing.

Just my mid-morning 2 Cents. :)

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons