There is an old idea kicking around that galaxies must interact strongly with the inter-galactic medium, especially when they fall into rich clusters. This idea comes from theoretical considerations (of, say, ram pressure), from observational considerations (of, say, galaxies clearly distorted by IGM interactions, and from the simple fact that galaxies in clusters have lower star-formation rates. This idea is wrong!
Actually, it is not totally wrong, as I discussed at length with Schiminovich today. There certainly are examples of galaxies that are interacting strongly with their IGM. However, it is not a dominant process in galaxy formation and evolution. No statistical observations of galaxy star-formation rates or activity as a function of clustocentric distance show the virial radius (or any other radius) strongly, and (with minor exceptions) galaxies do not
point the way to the centers of their nearest clusters. Most of these things I know from unpublished work (though some of it is here and here), but perhaps my long argument with Schiminovich today will convince me to dust some of that unpublished stuff off.