Skip to content

Latest commit

 

History

History
80 lines (63 loc) · 2.43 KB

README.md

File metadata and controls

80 lines (63 loc) · 2.43 KB

!Bin Chilling: A Metagenomic Ensemble Method

Docker file is also avaiable, and is the prefered way to run Bin Chilling. only tested on linux.

Quickstart guide:

  1. python3 setup.py
  2. python3 BinChilling -h
  3. follow guide

!Data requirements:

The data requirements of BinChillign are the following:

  • Fasta file
  • Depth/abundance file (when standable mode)
  • marker gene stats
    • file MUST end in .tsv
    • alternativly, a .scg file can be used, with format "EDGE_NAME\tSCG"
  • input partitions (unless in standalone mode)

!Docker Example: Data used in this example is the strong100 dataset, which can be found at: https://zenodo.org/record/6122610 Or can be directly downloaded from: https://zenodo.org/record/6122610/files/strong100.zip?download=1

Example assumes the contents of the dataset is located in the Data folder.

Add your dataset files to run following command in folder with dockerfile:

sudo docker build -t bin_chilling --no-cache .

Run docker container as using:

sudo docker run -it --rm -v /YOURPATH/P6/Data/:/P6/Data/ bin_chilling

This should start a docker container in interactive mode. Navigate to inside the BinChilling folder (Must be inside this folder, as the internal code expects the origin to be there.)

cd ./BinChilling

A help menu can be displayed using:

python3 . -h

Running in Ensemble mode:

python3 . -M Ensemble -f ../Data/edges.fasta -p ../Data/Partitions/ -g ../Data/marker_gene_stats.tsv -o ../Data/ens_output.tsv -j ../Data/edges_depth.txt --log ../Data/log.txt -ch 50 -am 0.90

Running in standalone mode:

python3 . -M Bin -f ../Data/edges.fasta -p ../Data/ClusterData/ -g ../Data/marker_gene_stats.tsv -o ../Data/bin_output.tsv -j ../Data/edges_depth.txt --log ../Data/log.txt -ch 50 -am 0.90

A file matching the output partition path file alongside one appended with .ref.tsv, will be outputted. The .ref.tsv file is the refined version, which is the optimized output.

-f = fasta file
-p = partition path
-g = marker genes
-o = output partition path
-j = depth file
--log = log file path (can be omitted)
-ch = chunksize (performance impact, does not change result)
-am = a1min value

!Evaluation Evaluation is done using marker_gene_stats the formulas from DAS_TOOls paper. Navigate to the scripts folder. python3 EvaluateBins.py -f ../Data/edges.fasta -g ../Data/marker_gene_stats.tsv -p ../Data/Bin_output.tsv.ref.tsv -d ../Data/edges_depth.txt -o ../Data/evaluation.tsv