Can't group samples in StringTie for differential expression

Question

[CandiceChuDVM]

Hi all,

I have performed StringTie for individual samples and --merge for all samples by implementing the following commands:

$ stringtie $PATH1/sample1.aligns.sorted.bam -G $ANNOT -p 20 -o $PATH2/sample1.gtf     
$ stringtie --merge -G $ANNOT stringtie_gtf_list -o $PATH2/merged.gtf

Now I am going to run StringTie using the -B/-b and -e options for each sample in order to estimate transcript abundances and generate read coverage tables for Ballgown.

However, I am not sure how to group my samples. For example, I would like to group all 9 samples into 3 different groups:

group1: sample1, sample2, sample3
group2: sample4, sample5, sample6
group3: sample7, sample8, sample9

Do I have to execute stringtie three times (one group at a time)? For example:

$ stringtie $PATH1/sample1.aligns.sorted.bam,$PATH1/sample2.aligns.sorted.bam,$PATH1/sample3.aligns.sorted.bam \
 -G $PATH2/merged.gtf -B -e -o group1.stringtie.out.gtf

Or can I run them all together by separating them with different labels as in Cuffdiff?

Thanks!