Alright you made me curious.
Paper based on looking at convicts.
From the extract: "Previous investigations have indicated that the ratio of sex offenders against female children vs. offenders against male children is approximately 2:1, while the ratio of gynephiles to androphiles among the general population is approximately 20:1."
"the ratio of heterosexual to homosexual pedophiles was calculated to be approximately 11:1. This suggests that the resulting proportion of true pedophiles among persons with a homosexual erotic development is greater than that in persons who develop heterosexually. "
Another paper that is even worse.
Article that goes the other way. It's based on looking at victims though, not at convicts.
The children in the group studied were unlikely to have been molested by identifiably gay or lesbian people.
pubmed.ncbi.nlm.nih.gov
These are all from the 90's, probably around the time that the gay marriage issue was starting be bandied about, so I'm not sure how much you can claim a lack of bias. Certainly can't claim a lack of any modern papers... if they existed.