ClonalFrameML for analyses of gene families about clonalframeml

[xavierdidelot]

Dear Julia,

ClonalFrame and ClonalFrameML have the same requirements in terms of input data, so there is no reason to use ClonalFrame rather than ClonalFrameML.

The best input is a whole-genome alignment as a multi-fasta file. You can also use a gene-by-gene alignment of core genes as a xmfa file, but this is not as good as a whole-genome alignement firstly because it includes less sites and secondly because you lose the information on gene order along the genomes. The xmfa format is not specific to Mauve, and is easy to generate for example if you have alignments of each locus then you can combine them into a xmfa file simply by adding '=' symbols between the loci.

You can also use a xmfa file with any number of loci (eg 7 for MLST) but the less genes are used the more difficult it is to correctly infer the clonal genealogy. If you have only a single loci for example then any recombination event replacing the loci will erase all information about clonal relationships.

Best wishes,
Xavier

[juliaped]

Thanks a lot, I really appreciate it! Now I understand.

Best regards,
Julia