Project Discovery is almost out (thanks to the recent patch delay) and is currently available for testing on SiSi. We previously gave a brief introduction to Project Discovery. So what’s the incentive?
Yeah, that’s right: you can earn lab-coats, battle armor and space drugs (I mean boosters…legal boosters). Analyst Kredits (AK) are the new operant conditioning method currency awarded for participation and being slick with Project Discovery. Below are the stats of the new, penalty-free. (Editor’s Note: These figures cannot be guaranteed to be accurate for the final patch.)
Slot 1
Antipharmakon Aeolis – 8% capacitor capacity bonus
Antipharmakon Kosybo – 8% armor repair bonus
Antipharmakon Thureo – 8% shield boost bonus
Slot 2
Antipharmakon Iokira – 8% tracking speed bonus
Antipharmakon Toxot – 8% optimal range bonus
So the 8% penalty-free boosters seem pretty cool, and yeah, that lab coat is mighty-space fancy. So what do you have to do?
In Project Discovery, you are tasked with identifying protein subcellular localization patterns (previously introduced here) for the Sisters of EVE. If you run through a whole bunch of slides, you will find yourself with news boosters and clothes in no time.
GETTING STARTED IN SCIENCE
The Project Discovery module is easy to find: just click on your Neocom in the top left, select ‘business’, and at the bottom you will find Project Discovery, which has a DNA-helix symbol. You can drag this onto your main bar if you click by clicking and dragging it to the left. I find this helpful for Planetary Interaction and now Project Discovery so I don’t need to go digging through the Neocom. Hell, it might just be sitting on the bottom of the left bar come patch day—that’s how it was on the test server.
Getting into the minigame is easy. Project Discovery is almost like a basic card matching game. I think of it as cellular mahjong, but with 30 choices to any sample. On the left you have the ‘Foreign Cell Sample’ which are always stained blue for the nucleus, red for the presumed cytoskeleton (probably against a protein known as α-tubulin, a component of the cell cytoskeleton), and the green is always the experimental protein staining. Putting your mouse over this image provides a really nice and responsive zoom feature. On the right side you have your options and oh-boy, my man, you got some options.
BYPASSING ALL THAT UNDERGRADUATE PLEBIAN WORK
To get you grounded, the game starts you out with a ten-slide tutorial with some easy matching options that are assisted. For successful matches, Professor Lundberg will doll out Analyst XP, a small ISK reward, and more often than not, Analyst Kredits. If your accuracy rating drops too low, then the rewards will also drop. If your rating goes up, the rewards get better. Got it? Good.
So what are all these options? Can I just choose nucleus or cytoplasm for everything? Maybe. To a well-studied molecular or cell biologist (or someone decently exposed to modern biology), the options on the right will make sense, and you will just be simply looking for the correct one. If you aren’t currently taking undergrad cellular biology or have a doctorate in molecular and cell biology with a few publications under your belt, you might be confused at first, but it’s easy to catch on.
You can look back and forth between at your pleasure, and several examples of each localization signature are provided for each based upon different levels of detection. What convolutes the analysis is it is very rare for a protein to have only one function or localization under any one specific cellular-state. Thus, you might have to choose multiple options in order to properly characterize the proteins localization status. This part will be tough for a novice. For instance, I have see vesicles (that exist in the cytoplasm) in several images, but it seems like I am the first one to see them in a lot preliminary try where others just say cytoplasm.
If you’re interested in digging deeper into the science behind Project Discovery, Wikipedia has an outline on all the different cellular structures you might find. Your options include the nucleus, nucleoplasm, nuclear bodies (few and many), nuclear speckles, nucleoli (rim, fibrillary center), cytoplasm, aggresome, mitochondria, rods and rings, cytoskeleton (intermediate filaments, microtubule ends, microtubules, cytokinetic bridge), microtubule organizing center, centrosome, endoplasmic reticulum, Golgi apparatus, vesicles, cell junctions, focal adhesions, and plasma membrane. Don’t worry. Even a seasoned cellular microscopist will find this list challenging.
After your tutorial, you will level up to gain more rewards and progress in your analytics career for the Sisters of EVE. You will often hit experimental slides which are only based on community consensus alone and have not been characterized yet. These seem to always give AK, and the answer is more diffuse and not specific (hence basing the final result on the consensus of players).
Making a consensus
ARE THERE PITFALLS?
Are there never not? While a major strength of Project Discovery is that all protein identities are blinded to the user, the lack of a color-blind mode (or greyscale option) leaves an entire population of people restricted from participating. Deuteranopia (red-green colorblindness) is found in at least 5% of the male population (.35% of females) by one estimate. The option of a greyscale or a color-blind mode (maybe white, cyan and teal instead of blue, red and green?) would allow many others to participate. Making fluorescent cellular images legible to all peoples is basically a requirement in any serious cell biology publication today. CCP should compromise, it is only logical given the state of the field.
One thing I always wanted to do while running through my test of Project Discovery was adjust the Z-axis of the image. If you draw a box on a paper that’s the XY-plane. Now imagine that box was 3D, what gives it depth is the Z-axis. Right now you are only given a single 2D optical section (of some thickness). However, cells do have depth, and the depth of the imaging focal plane can bias an interpretation. Maybe this is built in though, like it makes sense to only be judging one optical plane at a time, maybe on another example they show the same cells at a higher or lower point on the Z-axis. Either way, I have seen examples of differential depth between images. My guess is this is Z-bias is built into the statistical assembly of the final data somehow.
As a source for study, the game in its current form does not really do much to incite much thinking. It is mostly mindless (although addicting), and you get used to it rather quickly. With practice, you can get really good at identifying molecule localization patterns.
Unfortunately, despite your prowess, you are never told what the molecules are that you are looking at.
You will definitely learn some basic cell biology, but it will be completely devoid of significance or context. This is just probably my prejudice shining through. The information a cellular biologist might want anyways will be made available online at some point (out of EVE), as the results of the study are updated on a presumed continuous basis. It would be cool to be able to bounce back and forth on this within the game, but it makes sense not to have it included within EVE.
Because Project Discovery operates on a percent consensus, it would be interesting to see if the designers of the project have employed some sort of statistical test to correct for false negatives and false postives. Given I could see people just picking nucleus or cytoplasm as cop-out answers, being able to correct for this will be really important in deriving meaningful results. Even more interesting will be the downstream experiments performed based upon information gleaned from Project Discovery. You might just help cure some disease here.
GET YOUR SCIENCE ON
What we have here is a pretty cool little cell science simulator. The rewards seem worth it, although with the release delayed I am not sure how long it will take to acquire enough AK to get some boosters, a lab coat, or some battle armor. You get to learn and see different ways proteins can localize within these cultured cells. Hopefully Project Discovery opens some doors and minds for people not only interested in this line of work, but also to people that are just curious as to what we are made up of.
We are all just cells and stardust anyways, right?
