After KEGG and GO analysis, how to make tables+phylogenetic trees

Question

hope everyone is ok. I used Trinity to do a de novo transcriptome assembly, then blastp/blastx and then used Blast2GO software to do KEGG and GO analysis. So i got some txt files with header :

for GO

Sequence name Sequence desc. Sequence length Hit desc. Hit ACC E-Value Similarity Bit-Score Alignment length Positives

for KEGG

Seq Pathway Pathway ID #Enzs in Pathway Enzyme

All well so far. Then i was asked to do 3 tasks.

1. Tables with terpene synthases, (monoterpenes, sesquiterpenes, diterpenes and triterpenes) and corresponding phylogenetic trees including other known terpene synthases from all categories from other plant species.

2. Tables with Cytochrome P450s enzymes implicated in terpene biosynthesis (categorization per clades and clans) and corresponding phylogenetic trees including other known terpene CYPs from all clades and clans from other plant species.

3. From the graphs corresponding to the different KEGG and GO ontology analyses, export the data for the GO analysis for all three parameters (categories, names, blasts, total number of contigs and percentage) to an excel file to be able to transform them in other formats for future publication.

I have no idea how to proceed. Any guidance? Thanks

Answer 1

I'm going to answer 1+2 because 3 is hard to understand and maybe trivial?

Each of 1+2 seems to be have these sub-tasks:

1. identify the contigs in your assembly that have the genes of interest.
2. pull out the gene sequences from the contigs (you should have intermediate tblastx output somewhere with coordinates, if you don't you can just re-blast those contigs).
3. possibly translate those in silico to protein sequences.
4. tblastx (gene) or blastp (protein) those gene/protein sequences against whatever db is relevant.
5. pull sequences for the hits above some threshold for each gene from the db.
6. perform a multiple sequence alignment for each gene against the hits.
7. plug the MSA into whatever phylogenetic tool you like. I used phyml a lot back in the day, don't know what the kids like now.
8. make a visualization of the tree, probably?