The tiny bit of green is Native American or East Asian of some sort, from a reference of residents in Beijing and Tokyo. I got myself tested because I wanted to be surprised. This was a secondary test that I did, using a program from Stanford on a site called Interpretome (http://esquilax.stanford.edu/). Although it is only a tiny bit, it's exciting. Even though it is a tiny bit, it almost feels like a whole in a way, because without that one whole person, and that one whole lineage, I wouldn't exist. So I feel a new certain kinship. I guess I know it comes down to that we are all related, but sometimes it's harder to visualize and fit into relatable models for recent. Having this proof, makes this easier to visualize, and accept. The red-painted parts are all European. Green is East Asian, but Native American populations are the closest to that reference.
I recommend anyone curious to get tested (anyone really). 23andme.com doesn't offer health related reports anymore-- Only ancestry. But they give you access to your raw genetic data, which you can analyze using various other sites. I began by using http://SNPedia.com/, on my own, and referencing the data, but the programs are helpful, and some like http://geneticgenie.org focus on, and point out certain mutations that you have, that might give trouble. I found out a lot about myself in just a few days. Things that give me insight on my personality, social life, and how I naturally seem to handle stress, among the many things. Not to say I'm locked into this, necessarily, as there are ways to help myself.
I imagine this isn't totally accurate... 23andme's painting shows the East Asian/Native ancestry on chromosome 7 only. Perhaps they are picking up each different details, and they are on multiple. That would make more sense. I'm very new at this. I think I may send in another sample to another place to compare that data as well, as I wonder how accurate these test results are... If they would they pick up on the same results each time per person or if there are errors, and two tests from one might show slightly different results. It could be noise... I'd say, but multiple programs pick up on it.