Could I map strand specific RNA-seq with non-strand specific?

Question

[bxia]

I have some RNA-seq samples, some are strand specific, some are not.

Cuffdiff only can use either strand specific or non-strand specific, if I want to run all these samples together, could I just use the non-strand specific for cuffdiff? Except for the antisense RNA, is there a huge problem of doing that?

Thanks,

Answer 1

[Renesh]

It is not a good idea to map strand-specific RNAs as non-strand specific. Strand-specific RNAs has the criteria that first sequenced read will map to transcript strand and the other (in case of paired-end) map to opposite strand (for forward first strand protocol). For example, if you are using TopHat for mapping strand-specific reads to the genome, it will enforce this rule and will affect your mapping rate if you use library-type as non-strand specific. Ultimately, it will affect your cuffdiff analysis.