Processing TOGA Results to a Mutiple-species Orthogroups Table

[swagttt1]

Dear Professor Hiller,

@MichaelHiller

I hope this message finds you well. Since our last conversation, I have been attempting to resolve this #issue193, but it has been bothering me for several days, and I must reach out for your help again.

Assuming I have successfully run TOGA and obtained multiple results comparing reference species to query species (species1, species2, species3, etc.), each result contains several tables, such as orthology_classification.tsv, orthology_scores.tsv, loss_summ_data.tsv, etc. What I am aiming for is a table containing all species' orthologous gene groups (Orthogroups.tsv), similar to the output of Orthofinder or OrthoMCL. This would enable me to conduct subsequent convergent evolution analyses, among other tasks.

However, I am confused about how to process the results from TOGA (for example, how to handle the scores and loss issues for individual orthologous groups), and I’m unsure how to integrate the results from multiple reference vs query species into a single Orthogroups table containing all species.

Recently, I have gone through several of your lab's published papers, such as B10k and others using TOGA in the ZOONOMIA special issue, but I couldn’t find a readily available solution.

I was wondering if you have any pre-existing code or scripts that could help me generate the orthogroups table and convert these tables into multi-sequence FASTA files (OGxxxx.fasta). Alternatively, if you have a mature approach that I could follow to write my own script for extracting this information, I would greatly appreciate it.

I believe this issue could be helpful to many researchers, especially as more and more genomes become available, yet their gene structure annotations often remain lacking. By using TOGA, we could potentially achieve similar functionality to tools like Orthofinder.

I would be very grateful for your help.

[MichaelHiller]

Hi,

thanks for these questions. Always good to hear what people need.

Regarding the 'multiple species alignments' which we call 'multiple codon alignments', pls use extract_codon_alignment.py which is in the TOGA/supply directory. This performs exon-by-exon alignments and concatenates the results into a whole transcript alignments.

For this, you typically want to use 1:1 orthologs or allow a few species to have 2 co-orthologs. I think the script has parameters to enable this.

The orthology files for several query species can of course be integrated into a matrix of orthologs/orthogroups. Pulling out 1:1s is of course easy. For many:many orthologs, one needs to extract the whole orthogroup. We are working on a script to do that and will release this with TOGA2 (planned for Jan).

Also,

[swagttt1]

I know I ask a lot of questions, but thank you for your patient help.
Best wishes!