Masking A/C/G/T with N?

[Makoe]

I'm working with a tetraploid soybean species, and I want to determine which DNA-seq reads exist on only one of the two diploid progenitor genomes. I have two consensus fasta files (one for each diploid genome) and want to map reads to them to determine which reads map to only one genome based on mapping scores. I already have a pipeline for accomplishing this, but I'm running into a problem. One consensus file has 30 million more Ns than the other. Reads that should map equally to both species favor one because it has fewer Ns, and therefore a higher mapping score. Is there a way to replace A/C/G/T's in one consensus file with an N where the other consensus file has an N? The consensus sequences were made by mapping RNA-seq and DNA-seq reads to the Glycine max genome, so they should be comparable.

[N311V]

I'm not aware of a ready-made tool to do this job but it sounds like something you could do with Perl. I'm sure there are online tutorials you could use otherwise I found this book useful when first starting (http://shop.oreilly.com/product/9780596000806.do)

Are both consensus sequences the same length for each chromosome/scaffold? Doesn't really matter if they aren't, would just make the perl script to do this job easier.

[Makoe]

Ah, I guess I'll just have to build up my Perl skills. Probably a good idea, anyway. And yes, the consensus sequences are the same size, so it shouldn't be too difficult.

Thanks for your help!