I worked on a big report with a bunch of people (published online as an agency report). Can I extract my data and analysis from the report, expand on it, exclude what everyone else did, and publish separately in a journal? One of the Co-authors was also the Co-PI and did a lot of editing, plus brought in funds but I feel that co-pI did not contribute substantially to what I will extract for my journal article.
Usually large collaborations have some ground rules for who can publish on what with who as the authors. Make sure that you're not bound by those rules first.
Second, think whether you have anything to gain. Is this second paper going to really further your career if it damages your reputation among this group? What if none of these people or their postdocs and grad students ever want to Co-PI with you again? If you're asking whether your behavior is OK then you must have some feeling that it is not.
The only reasonable thing to do is to ask all the people involved. It's not like you're risking them learning something about your project that you want to keep secret because you're already collaborating. The standard thing is to contact them with your paper idea and ask whether they think they contributed substantially and want to be a co-author.