Open andersgs opened 7 years ago
Thanks Anders. Do you have an example?
Roary
doesn't actually add that Alg.-Conn
column - the header is a misnomer.
Fripan
takes input in the format of the output from proteinortho
, and that Alg.-Conn.
column represented "the algebraic connectivity of the respective group to indicate how densely the genes are connected in the orthology graph that was used for clustering". roary2fripan
simply reformats the Roary
output (gene_presence_absence.csv
) to resemble the proteinortho
output, so have just used a column that has some value in it.
My tests still seem to be working on the examples I'm using ...
Are you running roary with -s
switch?
"don't split paralogs"
This can cause problems, as you get >1 locustag in a cell.
@andersgs Do you know if this is still an issue? I tried running this recently on a few examples and didn't seem to have any problems.
Fripan is a total mess right now, sorry. @drpowell
In
proteinortho
file, Roary now adds aAlg.-Conn.
column. It seems to have every gene in the DB. It screws up the pan genome tree, the clustering, and the alignment views.Deleting that column seems to make things work again.
Anders and Jean.