Kalmbach Publishing has a book, The Model Railroader's Guide to Bridges, Trestles, and Tunnels that appears to have an example of this in it. Just put "central valley skewed girder bridge" into Google and it pops up as the first answer, even telling you it is on Page 33. Click on it and it opens right up to there. 

Here is a link to Google that should take you right to Page 33 - https://books.google.com/books?id=Y0-vEVUXb3QC&pg=PA33&dq=central+valley+skewed+girder+bridge&hl=en&sa=X&ved=0ahUKEwiCor-A4cDUAhVjHGMKHYNtAWQQ6AEIJjAA#v=onepage&q=central%20valley%20skewed%20girder%20bridge&f=false

I've skewed that bridge, and as it turns out, exactly as shown in the article that Jeff linked above - i.e., I moved one girder back to the next floor beam.  If it's skewed such that the floor beam doesn't make it to the other girder before intersecting the abutment, the beam would simply terminate on the abutment, probably on it's own plate or shoe. 

Here are a few sketches of skewed through truss bridges including situations were the skew is not a panel length:

http://sdrc.lib.uiowa.edu/eng/bridges/brenginvi/display.asp?p=485
http://sdrc.lib.uiowa.edu/eng/bridges/brenginvi/display.asp?p=487
http://sdrc.lib.uiowa.edu/eng/bridges/brenginvi/display.asp?p=486

The sketches are from J.A.L.Waddell Bridge engineering: http://sdrc.lib.uiowa.edu/eng/bridges/brenginvi/figures.htm

On skewed railroad truss bridge the stringers usually don't end on the abutment but in a floor beam as on o not skewed bridge. Only that this end floor beam is not perpendicular to the trusses.

Through truss bridges need portals at the ends to get lateral forces from the top bracing into the abutment. The end floor beam is part of this portal.
Regards, Volker

One trick I used with the Central Valley truss bridge on the Siskiyou Line is I cut the major-cross-beam panel lengths down from four sub-panels to three. I also shortened the bridge height somewhat to compensate for the shorter distance between verticals.

I occurs to me that shortening the floor by one sub-panel could make doing a skewed truss easier since the skew could be less.

Thanks, guys.  I figured that since it is fairly common on the prototype that someone had done a model version.

Robert