I'd just like to expand a bit on the minor citation: you often cite the seminal papers in a field simply to give context to the work you're doing and, in some fields, to show competing or past equations as contrast to your method of doing something.

None of those require the paper you're citing to be free from fraud: seminal works with fraudulent or inaccurate results may still have preparation methods that are applicable to novel, non-fraudulent, experimental techniques and equations are rarely at issue with data manipulation or other fraud.

There's still space to be aware of retracted papers in citations as building directly on those results and managing to show an improvement on doctored data would be suspicious but I'm pretty sure that is the minority of citations in many fields.