Open colbyga opened 6 years ago
Hi,
Should I submit my qc'd reads or refined bins as input to build a phylogenetic tree? As far as I can see I need to input my unassembled reads qc'd. Is this correct?
Do you mean to classify the reads or to build a gpkg? You can run graft
on reads or contigs/genomes, though you may want to input translated proteins instead to reduce noise, particularly since the ORF calling program OrfM is tuned for reads.
Is it possible to run graft on more than one "--graftm_package" at a time? eg. could I run it on all of the Ribosomal proteins in the list (https://data.ace.uq.edu.au/public/graftm/7/)?
Right now I'm afraid only a single gpkg can be run at a time, though several search HMMs can be applied at once.
Hope that helps, ben
Hi Ben, thanks for the advice!
I want to classify the reads/contigs and place them in a tree, I wasn't trying build a gpkg.
However, do you think it would be possible or ill advised to concatenate the 20 or so ribosomal fasta files and build a gpkg. Is there a reason not to do this? Is it because I wouldn't be able to have an alignment?
Right, the alignment at the junction of the proteins in the multiple sequence alignment might lead to artificial results. If you are interested in building a phylogenetic tree from ribosomal proteins, you might be interested in https://github.com/Ecogenomics/GTDBtk or if you want to link the contigs to reads using ribosomal proteins https://github.com/wwood/singlem
ben
Hi,
I am trying to use graftM for the first time. I have paired-end shotgun metagenomic data for 8 samples.
1) Should I submit my qc'd reads or refined bins as input to build a phylogenetic tree? As far as I can see I need to input my unassembled reads qc'd. Is this correct?
2) Is it possible to run graft on more than one "--graftm_package" at a time? eg. could I run it on all of the Ribosomal proteins in the list (https://data.ace.uq.edu.au/public/graftm/7/)?