There’s a solution, if some of your visitors are nerdy enough to do the test for you, you can find from where the call to the problematic content is coming.
I wrote this for my blog, I’ll copy-paste, just removing the name of my blog :
– first, open the blog and get the virus warning.
– then ask your browser to give you the source code of the website.
– then open your favorite text editor (not Word, a simple text manager, like Notepad or Notepad++ for MS-Winwows OSes), create a new text file, and paste inside it the code source you just obtained
– save that file as 1.html to a folder on your hard disk
– to make sure, open 1.html inside your browser : it should open a weird copy of the blog, and also trigger the virus warning
*** Now we’ll track which part of the source code calls for the virus ! ***
– with copy, cut and paste, create a new text file with the first half of the code of your 1.html file, and another new file with the second half of the code of 1.html. Save them as 2.html and 3.html
– Open 2.html in your browser, if it doesn’t trigger the virus warning, then it’s 3.html that should trigger it. If none of them triggers it, hit F5 with each of the pages a few times. You should find which of these two files will trigger the virus warning.
– And then again, you divide the culprit file in two halves (for instance 2a and 2b.html), and so on, and so on, until you have a very small html file responsible for the virus warning !
– ideally, that final virus-causing file should contain just a few lines of code, allowing to see precisely what item within the website is calling for a virus attack