Hi all,

While I'm doing a pangenome analysis for 2 different group of strains with Roary, I find out one problem on gene clustering.

What I want to do is to find differentially existing genes between the groups, and within group. So I run Roary two times, one for integrated strain data and one individually for either group (A+B groups strains, A/B group strains respectively). And I find out that gene clustering has been changed like below:

1) When run Roary on data of either group (A/B): Group A --> (S1, S1', S1'', S1''') are clustered together as "Gene k" Group B --> (T1, T1', T1'') are clustered together as "Gene j"

2) When run Roary on data of integrated group (A+B): (S1, S1', S1'', T1') are clustered together as "Gene k" (S1''', T1) are clustered together as "Gene M" (T1'') is clustered as "Gene j"

In this case, how can I make them have consistent clustered outcome??