The Qiime version is: q2cli version 2022.11.1 using Miniconda
I used QIIME2 in order to build a FeatureTable[Frequency] artifact. I exported it as a Biom table, which later I converted to a tsv table, and used it for running the FUNGuild script.
After exporting and converting the table, I formated it making sure that it corresponded to all criteria on FUNGuild's GitHub (deleted the first line, removed the # in front of OTUID, added header column taxonomy and, using Excel Sheets, checked that everything alligned accordingly), however upon calling the script I get "Found 0 matching taxonomy records in the database" as a result.
Could anyone please let me know if I made any mistakes while creating the table, if so, how could I fix them? Unfortunately the link to the OTU_table_example over at GitHub cannot be accessed anymore, so I only followed the written instructions. Or is there anything I might have done wrong while using QIIME2 to buid the table?
Hi @Riva ,
I have re-categorized this topic as "other bioinformatics tools" because it sounds more like an issue with FUNGuild, not with QIIME 2. Your feature table looks fine, so I see nothing wrong prior to exporting.
I have never used FUNGuild before, so cannot help on that end (you should post an issue in their issue tracker instead). But a possible lead: it looks like the taxonomy column in your 2_mort_otu_table.txt does not actually contain any taxonomy information, so I would guess that needs to be filled before passing to FUNGuild. Just a guess.
Thank you very much for your reply and time. And also for reassigning the topic to the right category.
About your suggestion regarding the taxonomy column., I thought that it is was meant to be left empty as the FunGuild script should be the one performing the assignment.
If this is not the case, then the table should be obtained by following this type of procedure right?
But in order to obtain a taxonomy artifact, a classifier needs to be trained and the used. However, in my case, since the data is fungal DNA, the database for training is UNITE, and I don't have the necessary resources to train it myself. I've searched for a pretrained calssifier, and this was the most recent version I could find. But it was made using QIIME2 2019.1 version. I've been trying to install that version in a seperate environment, but it hasn't woked out well.
So I when looking for ways to avoid this issue, I thought I could use FUNGuild to perform the assignment, therefore not needing to fill out the taxonomy column, just it having to be introduced as a mandatory header.
I am sorry to have poured all of this on you, especially since you mentioned you haven't worked with FUNGuild before. I just thought maybe someone would see this message and be able to reply.
You are probably right. I have no clue, as I have not used FUNGuild before. I thought that FUNGuild assigned taxa to "guilds" based on the pre-assigned taxonomy.
Thank you so much for your reply, Sir!
Regarding the taxonomy column, I think it's me who has misunderstood. I will continue by using the classifier you linked to fill out the column and try again after.