Short answer: Not that I have seen. Long answer:
This metaphor is a bit difficult to work with, because a family tree essentially shows two components that are necessary, sufficient, and unequivocally causal to producing the third: mother+father=child. I'll talk about the two major issues with such a model: complexity and context.
First, Wars.... aren't generally that simple and traceable. For example, if you wanted to look at JUST the wars that contributed to World War I, you'd need to understand the impacts of (in no particular order) the Crimean War, the Russo-Turkish war, the Franco-Prussian war, the Russo-Japanese war, the first and second Balkan wars of 1912, and the first and second Boer wars -- and that's just within the latter half of that century! You might expand this to show more than 2->1 relationships, but the late 1800's are still going to end up looking more like a web than a tree. This issue is mostly one of clarity. Namely, if you can't make it look like a tree... the model doesn't really do much for you.
The second issue is much more serious: context.
Put simply, while wars certainly inter-relate, in almost all circumstances wars don't just contribute to the next conflict on their own. There are definitely circumstances in which two wars clearly relate (the two World Wars being a well-known example), but it is a drastic over-simplification to say that "War X caused War Y." There are dozens of concurrent trends in culture, science, technology, philosophy, and politics that weave into the mix.
Let's go back to World War 1 and its "tree." You could draw out a very large web of all the late 1800's conflicts that led to the web of alliances in Europe, but missing from that would be so many vital factors: the rise of Nationalism and the redefining of national identities (especially in the Balkans), or the impact of Imperialism on how people thought about conflict (or resources), or the massive technological and industrial changes that were continuing to fundamentally change the structure of societies (as well as their machines of war).
My point being that such a tree would be, at best, a timeline of conflicts sorted by geography and lacking context. At worst, it presents history in a drastically oversimplified format.
I haven't come across such a thing anywhere, but I would love to see such an event tree mapped out. With the rise of more and more sophisticated graphing and imaging software, an interactive, zoomable interface tracing the lines of connection between global conflicts throughout history could be astoundingly useful and fascinating.
Here's how one might start: make an exhaustive list of global conflicts, along with all dates and locations involved. Theoretically, one could create a three-dimensional "tree" of any given conflict using only the geotemporal data, using the three axes of time, latitude, and longitude. One could then link conflicts together from there, and refine the structure with color coding denoting battles, treaties, ceasefires, with webs of diplomacy linking distant locations, etc.
There used to be a website called "Conflict History" which mapped out conflicts by year and location, but it's since stopped working, unfortunately. http://www.mapsofwar.com/ has been around for a long time and might provide some inspiration as well.
A good place to start, just to get a handle on the sheer amount of information involved, might be compiling and/or building visual timelines, in the manner of this one, or, on the more complicated end of the spectrum, this one from The Economist, as well as those found on hyperhistory.com as well as here.
You might also be interested in the Grand Unified Timeline of Human History project.
At any rate, it sounds like it would be an insanely ambitious endeavor, but one that could be immensely rewarding.
EDIT: If you like data and charts, The Center For Systemic Peace's conflict trends page is a great resource. I also found this fun video attempting to visualize world conflict over the past millennium.
