Installation: make sure you have all required external softwares installed, and fill the path to the relevant executables in: GutReferenceSet/Utils/config.py
fill in the working directory (or the Demo directory) in: GutReferenceSet/Build_Species_Set/config.py
if trying the Demo - correct the index of GutReferenceSet/Demo/full_metadata.csv to the full path of your system
Run and tested on: python 3.7.4 on CentOS Linux 7.9
Creating a new reference set:
-
Create a list of files and their metadata A file named full_metadata.csv with index witch is the full path of an assembled genome fasta file and at least the following columns: Source, Method, AssemblyName, SampleName, RegistrationCode, DoNotTake Where:? Source - source of data, so that Source + RegistrationCode is a unique identifier of an individual from which assemblies where created Method - assembly creation method (MAG / isolate / nanopore...) AssemblyName - a unique identifier of the assembly SampleName - an identifier of the sample the assembly was created from (so as to identify assemblies originating from the same sample) RegistrationCode - an identifier of the individual the sample was taken from (so as to identify assemblies originating from the same individual) DoNotTake - a columns which is either empty or includes reason not to consider the assembly
-
Calculate the qualities of the genomes by running checkM: GutReferenceSet/Build_Species_Set/quality.py
-
Filter by quality and by criteria of not taking same species from same person twice: GutReferenceSet/Build_Species_Set/filter.py
-
Mash all vs all into a memory mapped file using: GutReferenceSet/Build_Our_Set/distance.py
-
Cluster based of memory mapped distances. Using: GutReferenceSet/Build_Species_Set/hierarchical_clustering.py
-
Choose representatives: GutReferenceSet/Build_Species_Set/choose_representatives.py
-
Name representatives with GTDB: GutReferenceSet/Build_Species_Set/naming.py
-
Compare to Segata and build tree structure: GutReferenceSet/Utils/phyphlan.py