Contact

Dr. Shan Zhang (link)¹ and Dr. Weizhi Song (link)²

¹ Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong

² Department of Ocean Science, Hong Kong University of Science and Technology, Hong Kong

Email: shanbio@hku.hk, ocessongwz@ust.hk

What has been changed:

1. The input genomes to MetaCHIP2 must be in GenBank format. If your genomes are currently in FASTA format, you'll need to perform an initial annotation step before feeding them to MetaCHIP2. You can use MetaCHIP2 prokka module to batch generate the .gbk files for your input genomes.This pre-annotation strategy could:

   • bypass the need for repeated genome annotation when exploring MetaCHIP2 parameters, thereby reducing computational time.
   • minimize the introduction of variations from the annotation process itself, and thus
   • ensure better comparability of predictions between independent MetaCHIP2 runs (on the same set of input genomes).

2. The user now need to provide a species tree for the input genome. Again, this could avoid repeated tree inference, which in turn leads to more consistent and comparable predictions between separate MetaCHIP2 runs on the same set of input genomes. You can use MetaCHIP2 tree module to infer the species tree, This module wraps GTDB-Tk's identify, align, and infer functionalities.

3. The inferred species tree must be rooted, as required by Ranger-DTL (one of MetaCHIP2's dependency). If you use MetaCHIP2 tree module for tree inference, the tree will be automatically rooted according to the GTDB taxonomy. If you use your own way to get the species tree, please make sure that it is properly rooted.

4. The PI and BP modules in MetaCHIP has now been merged into a single module called detect in MetaCHIP2.

5. You can now use mmseqs linclust (by specifying '-m' to the MetaCHIP2 detect module) to speed up the time-consuming all-vs-all blastn step in MetaCHIP2.

6. The output files are now organized in a more intuitively way, making them easier to understand.

7. More detailed interpretation of the donor gene/genome, and the often-observed low similarities between the donor and recipient genes. Please see details below in the "Output files" section.

8. A changelog is here.

Dependencies:

• Python libraries: BioPython, Numpy, SciPy, Matplotlib and ETE3.

• Third-party software: GTDB-Tk, BLAST+, MMseqs2, MAFFT, Ranger-DTL 2.0 and FastTree.

Install MetaCHIP2 with Conda:

• As MetaCHIP2 requires GTDB-Tk, we'll create a Conda environment pre-installed with GTDB-Tk. You'll need to setup the database files for GTDB-Tk as described in its manual.

  conda create -n metachip2env -c conda-forge -c bioconda gtdbtk=2.7.1
  conda activate metachip2env
  pip install MetaCHIP2
  conda install -c bioconda blast
  conda install -c bioconda mafft
  conda install -c bioconda mmseqs2
  conda install -c bioconda diamond
  conda install -c conda-forge r-base
  conda install -c conda-forge legacy-cgi
  

• Upgrade with: pip3 install --upgrade MetaCHIP2

How to run:

• The input files for MetaCHIP2 include a folder that holds the gbk file of all query genomes, as well as a text file which provides taxonomic classification (example) or customized grouping (example) of the input genomes. File extension (e.g., gbk) of the input genomes should NOT be included in the taxonomy or grouping file.

• GTDB-Tk is recommended for taxonomic classification of input genomes. Only the first two columns (user_genome and classification) are needed.

• Input files for MetaCHIP2 must be in GenBank format. You can run MetaCHIP2 prokka -h to batch generate the .gbk files for your input genomes. To prevent potential Prokka errors, please ensure that contig IDs remain shorter than 18 characters.

• The user now need to provide a species tree for the input genome. You can run MetaCHIP2 tree -h to infer the species tree, which wraps GTDB-Tk's identify, align, and infer functionalities. The inferred species tree must be rooted, as required by Ranger-DTL (one of MetaCHIP2's dependency). If you use MetaCHIP2 tree for tree inference, the tree is automatically rooted according to the GTDB taxonomy. If all genomes on the species tree are from the same genus, MetaCHIP2 tree will root it at middle point. If you use your own way to get the species tree, please make sure that it is properly rooted.

• Now you are ready to detect HGTs among your input genomes.

  MetaCHIP2 detect -i gbk_dir -x gbk -c taxon.tsv -s rooted.tree -t 12 -f -o op_dir -r pcofg
  

• You can use mmseqs linclust (by specifying '-m' to detect module) to speed up the time-consuming all-vs-all blastn step.

  MetaCHIP2 detect -i gbk_dir -x gbk -c taxon.tsv -s rooted.tree -t 12 -f -o op_dir -r pcofg -m
  

• If you already have the all-vs-all blastn results on the same set of input genomes from a previous run, you can skip the blastn by providing the blastn results with '-b'.

  MetaCHIP2 detect -i gbk_dir -x gbk -c taxon.tsv -s rooted.tree -t 12 -f -o op_dir -r p -b path/to/previous/run/blastn_op
  

• Options for argument '-r' can be any combinations of d (domain), p (phylum), c (class), o (order), f (family), g (genus) and s(species):

  MetaCHIP2 detect -i gbk_dir -x gbk -c taxon.tsv -s rooted.tree -t 12 -f -o op_dir -r pcofg
  MetaCHIP2 detect -i gbk_dir -x gbk -c taxon.tsv -s rooted.tree -t 12 -f -o op_dir -r pco
  MetaCHIP2 detect -i gbk_dir -x gbk -c taxon.tsv -s rooted.tree -t 12 -f -o op_dir -r ofg
  

Output files:

• A Tab delimited text file (detected_HGTs.txt) containing all identified HGTs.

  Column	Description
  Gene_1 [1]	The 1st gene involved in a HGT event
  Gene_2 [1]	The 2nd gene involved in a HGT event
  Identity [2]	Identity between Gene_1 and Gene_2
  Occurence(taxon_ranks)	Only for multiple-level detections. If you performed HGT detection at phylum, class and order levels, a number of "011" means current HGT was identified at class and order levels, but not phylum level.
  End_match	End match or not (see examples below)
  Full_length_match	Full length match or not (see examples below)
  Direction [3]	The direction of gene flow. Number in parenthesis refers to the percentage of this direction being observed if this HGT was detected at multiple ranks and different directions were provided by Ranger-DTL.

  ^[1] The most accurate interpretation of the "donor gene" is "the gene from the donor group of your input genomes that exhibits the highest similarity to the recipient gene".

  ^[2] A low similarity does not necessarily indicate it's an ancient gene transfer. Instead, it more likely reflects the absence of the exact donor organism (the organism that physically contributed the transferred gene) in the input genomes.

  ^[3] Similar to the interpretation in [1], the donor genome is the genome within the donor group that contains the gene exhibiting the highest similarity to the recipient gene.

• Nucleotide and amino acid sequences of identified HGTs.

• Flanking regions of identified HGTs. Genes encoded on the forward strand are displayed in light blue, and genes coded on the reverse strand are displayed in light green. The name of genes predicted to be HGT are highlighted in blue, large font with pairwise identity given in parentheses. Contig names are provided at the left bottom of the sequence tracks and numbers following the contig name refer to the distances between the gene subject to HGT and either the left or right end of the contig. Red bars show similarities of the matched regions between the contigs based on BLASTN results.

• Gene flow between groups. Bands connect donors and recipients, with the width of the band correlating to the number of HGTs and the colour corresponding to the donors, band arrow points to the recipient.

  If you want to visualize gene flow for a subset of detected HGTs (e.g., HGTs belong to a specific functional group), you can subset the "detected_HGTs.txt" to keep only the interested HGTs and run the circos module. The grouping file is in MetaCHIP2's output directory.

   MetaCHIP2 circos -l detected_HGTs_subset.txt -g grouping.txt -o interested_HGT_circos_plot.pdf
  

• Enrichment of COG functions in the detected HGTs (to produce a plot similar to Fig. 9 in the MetaCHIP paper)

   MetaCHIP2 enrich -f -diamond -t 12 -faa faa_dir -o op_dir -db path/to/COG_db -hgt1 detected_HGTs.faa
   MetaCHIP2 enrich -f -diamond -t 12 -faa faa_dir -o op_dir -db path/to/COG_db -hgt1 Setting1_HGTs.faa -hgt2 Setting2_HGTs.faa -label1 Setting1 -label2 Setting2