As a caveat, believe virtually no claims of a "gene for X" made in the popular press. Most traits are complex, involving many different genes and complicated interactions with the environment. Eye color, for example, has at least 3 genes affecting it, each with multiple alleles, and is not the simple 1 gene 3 allele model most of us learned in high school biology. Even traits where there's a single gene underlying it -- sickle cell, Huntington's, Phenylketonuria, etc. -- there is both an environmental component, and the "gene for X" is actually a gene that is supposed to do something else entirely, but when it malfunctions it leads to X. For example, people with PKU can't digest phenylalanine, which builds up to toxic levels in developing neurons. By having a restrictive diet which avoids phenylalanine through childhood and adolescence, virtually all of the mental problems associated with PKU are avoided. The PKU gene isn't there to give someone mental problems; it is a version of the gene to digest phenylalanine which doesn't work properly, which therefore can lead to mental problems.
From the scientific literature on homosexuality, the greatest statistical support is for an influence of maternal hormones. Specifically, the probability that a given male fetus will develop into a gay man increases with the number of previous male pregnancies that the gestational mother has carried, regardless of whether those males are alive or in the household where the baby that this particular fetus will become is raised. No such link has been found with probability that a given female fetus will develop into a lesbian. It is theorized that previous male pregnancies trigger a low grade immune reaction in the gestational mother toward testosterone and other androgens -- which end up in higher concentrations in a woman pregnant with a male than she'll have at other times in her life -- and that this immune response and/or her own endocrine reaction to the circulating male hormone levels ends up influencing the brain architecture of the fetus/embryo (depending on when in gestation it occurs).
This is among the best examples of how something can be inborn, biological, and not genetic. Both the genetic children of homosexual individuals, and the children adopted by homosexual individuals, are no more likely to be homosexual as adults than is the general population. The greatest statistically significant difference is in their self-identification as queer during their teens and early 20s, which is most likely the result of not staying in the closet for fear of parental rejection; by the mid 20s, the difference is no longer significant (last time I looked at the data).
A much weaker line of evidence points to the possibility of genetic influence of a particular region of the X chromosome, specifically regions Xq28. However, whether or not this region is statistically significant is borderline -- it's right on the standard threshold genetic researchers use for significance in searching for genetic regions influencing a trait. Interesting to note: this region is on the X chromosome. Thus, if it does have an effect, you would expect to have more gay relatives on your mother's side of the family than on your father's, as your X chromosome came from your mom, not your dad (assuming you are genetically male).
As for me: no known blood relatives are gay. The relatives I am aware of who are gay are all relatives by marriage -- one of my mother's cousin's brothers-in-law, for example. But, I am just one person, and thus an anecdote. To understand what's really going on, we need very large sample sizes, and ones free of bias from self-selection in reporting.