Phylogeny inference on complete chloroplast genomes with IQ-TREE software

Question

[Denis]

Hi,

I'd like to apply the IQ-TREE software for phylogenetic reconstruction on multiple nucleotide alignment of complete chloroplast genomes. Which parameters of the program should i use for that? The genome typically contains a lot of introns, CDSs, intergenic spacers and etc., so it would not be easy to create a numerous independent partitions for phylogenetic analyses. Are there any other appropriate solutions for IQ-TREE ? Which model should i choose?

Many thanks!

Answer 1

[shelkmike]

Some plastid genomes have rearrangements, so I wouldn't align them entirely. In my experience, it's better to do the following:
1) Align genes separately.
2) Remove poorly aligned regions using Gblocks.
3) Concatenate these alignments into one alignment by geneStitcher (https://github.com/ballesterus/Utensils). geneStitcher will produce a file with a list of partitions. After a slight modification, it is suitable for IQ-TREE.
4) Build a tree using IQ-TREE with options "-B 1000 -m MFP --merge".

The file with partitions that you provide to IQ-TREE should have lines like:
CODON, rps3_alignment_after_gblocks = 2704-3162
CODON, rps4_alignment_after_gblocks = 3163-3480
CODON, rps8_alignment_after_gblocks = 3481-3876
DNA, rrn16_alignment_after_gblocks = 3877-4989
DNA, rrn23_alignment_after_gblocks = 4990-7132
DNA, rrn5_alignment_after_gblocks = 7133-7237

Which means, you must indicate which genes are protein-coding and which genes are RNA-coding.

[Denis]

Thanks for your detailed reply! But how should i deal with genes containing introns in their sequences? Do they belong to the CODON or DNA type?

[shelkmike]

Use sequences without introns and UTRs. To get them, I use the program genbank_to_fasta (https://github.com/arundurvasula/FPS-repository/blob/master/genbank_to_fasta.py).