When building OrthoGroups and Gene families for CoNekT the easiest way to go about this is to add the species to the database, and export the protein fastas from CoNekT. This ensures all IDs are the same.
Output from OrthoFinder and (tribe)MCL can be directly imported, add a fitting description, select the type of data you wish to import and select the file. Hit Add Families to upload the file and create the gene families in the database.
For OrthoFinder, select Orthogroups.txt, from the output. For (tribe)MCL pick the file with the final output (all members of a gene family on one line).
OrthoFinder's phylogenetic trees can be imported into CoNekT. To do so first create a gzip file containing all the trees. Furthermore you will need to locate the file SequenceIDs.txt which is used to convert OrthoFinder's internal IDs back to CoNekT's.
First select the OrthoFinder families you wish to add trees to. Next add a description and finally select the gzip file with all trees and SequenceIDs.txt.
Currently adding trees to other types of gene families is not supported.
For clades to be detected correctly, clade definitions need to be stored in the database from 'Add->'Clades'. This is done using a JSON object structured as the example here:
{
"Arabidopsis": {
"species": ["ath"],
"tree": null
},
"Poplar": {
"species": ["ptr"],
"tree": null
},
"Rice": {
"species": ["osa"],
"tree": null
},
"Rosids": {
"species": ["ptr", "ath"],
"tree": "(ptr:0.01, ath:0.01);"
},
"Angiosperms": {
"species": ["ptr", "ath", "osa"],
"tree": "((ptr:0.03, ath:0.03):0.01, osa:0.04);"
}
}Dictionary keys are different clades, within each dict you have to specify two things : the species, which contain an array of short names of the species in that clade and a tree with a newick tree of that clade.