Single vs. multiple tophat runs

[gkcn]

Hello,

What is the difference between running:

tophat indexfile read_a_1,read_b_1 read_a_2,read_b_2

and running

tophat indexfile read_a_1 read_a_2; tophat indexfile read_b_1 read_b_2
consecutively?

Let's assume that read_a and read_b are different samples (biological replicates?) from the same cancer tissue and we're looking for some ncRNA regions. I mean we will not compare expression of certain genes in replicates to each other.

[Bukowski]

Nothing as far as I know, it is Cuffdiff handles the replicate groupings for comparison purposes.