Depends on what you are disrupting.
If you are disrupting how a particular piece of core software technology works - then Open Source, which requires you to republish the derivative work - makes such disruption harder. Think about what Tivo had to do to build its video player on Open Source. Essentially they had to build a generic video driver and storage using the open source stuff - and publish that back.
THEN they had to build a separate device driver for video and storage completely from scratch with a different team that did not know or could look at the Open Source source code. So that they could say "its not derivative so we don't have to republish"
OTOH, if you are disrupting something via business process- say a building the next Uber - which is basically just a automated dispatch system - then none of what you are writing is "derivative" and hence you don't owe republishing the code.
So if you are looking to patent something that is based on the Open Source technology it self, you cannot. That's the purpose of Open Source
IF otoh you are looking to patent say the idea of using big data analytics to determine what prices taxi passengers are willing to pay while keeping all your cars fully occupied, that you could use open source for and not have a problem with the patent since you are patenting your analytics, and not the code tool
It takes roughly the same amount of effort to write the code from a "closed source" API as from an open source one.
So if you want to leverage Open Source as your base say for the 25% of your idea that is disruptive here's what you need to do:
mostly this adds a bit of cost in the spec process and in the back and forth on the bugs in the 'black box". But its not prohibitive. You just have to be careful with the process.
you can always hire an outside consultant who has experience in doing this (like me :-) ) to help you in some of the crucial points