In general, the coverage sparsity will make this a pointless task. Many genes in scRNA-seq only have 1 or 2 reads (or none), and they'll be biased location-wise in this case. False negatives and positives will abound. I expect this will be a waste of your time. In fact, without WES/WGS, I don't know how you'd dissociate anything you find from actual SNPs. Sequencing errors will result in false positives.

Save yourself the headache and talk things over with your PI or colleagues to identify more robust approaches if this is really an avenue you'd like to pursue.

This paper covers some of the challenges of accurately detecting RNA editing events. And that's with deep transcriptional profiling.