When scaling recipes, I've always used the same approach as Jimmy.
No doubt this will, strictly, produce a slightly incorrect result because not everything scales up and down in a linear way. You get (in)efficiencies that remain constant (like mash tun deadspace) and efficiencies that scale (like the tendancy for bigger systems to generally be more efficient due to economies of scale). The first time I ever brewed anything, I made an AG version of Siren's Liquid Mistress, scaled from 6 gallons to 5 but because of wastage and being new to the whole experience, all I got in the bottle was about 3.8 gallons. It was a really good beer.
Bottom line, for your first AG batch, I wouldn't worry about getting it spot on to a huge degree of precision. If you have software that will give you a result, like Beersmith or BrewTarget, start by scaling down the inputs to 61% of what you would have used, then see if you get consistent ABV, FV/OG, SRM, IBU etc. If they're pretty similar, go ahead.
In the long run, chasing inefficiencies, scaling and modifying some recipes is great, but don't let that stop you experimenting at the start and learning what works/doesn't work through trial and error - it'll give you a much better appreciation of the whole process and systems that are interacting.
