Multigene-Phylogeny: How to merge several Multiple-Sequence-Alignments?

Question

[Tom.Ma]

Hello, i'd like to build a multi-gene-tree of 14 fungal genomes. I did so far:

• Analyze the protein-sequences of all genomes for shared orthologs using orthoMCL
• Filter the resulting groups/clusters for those containing (only) one ortholog of every genome
• Generate 299 separate Fasta-Files of those groups with the sequences of the orthologs
• Run a MAFFT-alignment on every one of them separately

My files look like this one:

Grph|08566 ----------------RGAFEWAGNSVGGLFCQASNIVSPSHWRMVWDVVRFNYQSIASLRAFDRASEEQ---------------------------- Rany|05125 ----------------MRIAVIGSGVSGLAATWALNEAGNSVGGLFCQASNIVAGNSVGGLFCQASNIV----------------------------------- Acai|269125 ----------------MRIEWAGNSVGGLFCQATWALNEAGNSVGGLFCQASNIVAGNSVGGLFCQASNIV----------------------------

The order of the species (e.g. Grph) in those files is varying from file to file.

My next step would be, to run RaxML.

My question is: How do I combine those 299 separated alignments to get one "multi-gene-alignment" as Input for RaxML?

Thank you for your help.

[liupfskygre]

you may try concat from seqkit.

[Tom.Ma]

Thats been a good tip, thanks!

[Joe]

You can merge the MSAs 'end to end', and then with RAxML use a partition file (I think). The partition file is needed to demarcate where each alignment starts and ends.

Another option might be to compute a consensus/species tree from all of the gene trees via ASTRAL or similar.

Answer 1

[Brice Sarver]

I provided this solution for another question, but it will work just as well for combining MSAs: A: merge two multifasta files