This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Multiple alignment of genome assemblies to generate a tree

I have 2000 genome assemblies(each fasta might have 5-6 contigs) of a particular serovar of Salmonella in 2000 separate fasta files(essentially 2000 fasta files). I need to align them to each other so that I can generate a tree. What would be the best tool to align these sequences and the aligned output file needs to be a single fasta file. Thank you.

genome-assemblies multiple-sequence-alignment
• 1,755 views • link
written by pramach1 • pramach1

Why do you need a tree? A tree with 2,000 tips isn't really good for anything and it would take pretty much forever to build..

I suggest all-vs-all Mash (choose k and s wisely) and subsequently cluster the resulting distance matrix with affinity propagation. Even with a laptop, this takes but a few minutes. If you use the R AP implementation, you can also output a heatmap with a dendrogram..

• 0 views • link
written by 5heikki 5heikki

Thank you. This answer led me to discover mashtree. I have access to a cluster and it was really fast. Thank you.

• 0 views • link
written by pramach1 • pramach1

1 answer

It is pretty much impossible, and for visual purposes intractable, to align 2000 full bacterial genomes. If you really want to stick with it, this program will make a cladogram based on average nucleotide identity:

https://github.com/MrOlm/drep

Beyond that, I suggest you remove all the genomes that are less than 20-30% complete, select a set of ~100 single-copy protein markers, and then make a tree based on concatenated protein alignments. That is going to be more tractable than actually making genome alignments.

• 0 views • link
written by Mensur Dlakic Mensur Dlakic

Log in to answer this question.

More posts like this