Inquiry on inferring directionality of HGT events in WAAFLE

I am writing to ask how WAAFLE infers the direction of HGT events. WAAFLE assigns directionality to syntenies such as ABA, AB~A, and ABA~A, but not to patterns such as ABABA, ABA*A, AB*AA, ABA~*A, or AB~*AA.

Could you clarify the rationale for this? I would have expected some of these patterns to be directional if direction is considered at the level of individual genes rather than the entire contig. For example:

  • ABABA could be interpreted as containing two B>A transfer events.

  • Because * denotes a gene assigned to both taxa, I would not consider it a transferred gene itself. I would therefore expect it not to affect the inferred direction of the adjacent B gene in patterns such as ABA*A, AB*AA, ABA~*A, or AB~*AA. This seems analogous to ignoring genes marked with ~.

For example:

Synteny Default direction Possible interpretation Proposed direction
ABABAA A?B ABABAA → ABA + ABAA B>A, B>A
ABAAA* A?B Ignore *, or assign * to A B>A
AB*AAA A?B Ignore *, or assign * to A B>A
AB~A B>A ABA after ignoring ~ B>A

I would be very interested to know whether this behaviour reflects a deliberate conservative rule in WAAFLE, and if so, what motivated that choice.
Thank you for your time and efforts.

1 Like