aaKomp

Description

Assess draft genome completeness using a fast, alignment-free, k-mer hash-based approach (aaKomp). This tool uses amino acid k-mers and a multi-index Bloom filter (miBf) to estimate the completeness of genome assemblies.

Credits

Concept: Johnathan Wong and Rene L. Warren
Design and Implementation: Johnathan Wong

---

Installing from Conda (Preferred method)

Under construction

Installing from Source

Clone from GitHub

git clone https://github.com/bcgsc/aakomp.git
cd aakomp
meson --prefix /path/to/install build
cd build
ninja install

---

Dependencies

• GCC 7+ with OpenMP
• Python 3.9+
• zlib
• meson
• ninja
• tcmalloc
• sdsl-lite
• libdivsufsort
• btllib
• libsequence
• gperftools
• boost-cpp
• r-base
• r-ggplot2
• r-dplyr
• r-readr
• r-cairo
• r-gridextra
• r-pracma
• hmmer=3.1
• pigz

Installing Dependencies with Conda

We recommend creating a fresh conda environment:

conda create --name aakomp
conda activate aakomp
conda install -c conda-forge -c bioconda --file requirements.txt

---

Running aaKomp

You can run aaKomp either directly or using the driver script run-aakomp.

Driver Script: run-aakomp

The run-aakomp driver automates:

• Downloading BUSCO lineages
• Building a miBf if missing using make_mibf with BUSCO lineages or provided references
• Running aakomp
• Visualizing with aakomp_plot.R

---

Demo Example

Here are two example usages of run-aakomp. In both cases, the --db-dir flag controls where the miBf (multi-index Bloom filter) is stored and looked up.

# Option 1: Run aaKomp using a provided reference file
run-aakomp --db-dir ./ \
  --reference reference.faa \
  --input input.fa \
  -t 4 \
  -o output_ref
  # --visualise optional argument to visualise the cumulative distribution function

# Option 2: Run aaKomp using a lineage name (e.g., "eukaryota")
# The lineage's HMMs will be downloaded and consensus sequences will be extracted to generate a reference
run-aakomp --db-dir ./ \
  --lineage eukaryota \
  --input input.fa \
  -t 4 \
  -o output_eukaryota

Note:
If the required miBF already exists in the specified --db-dir, it will be reused. Otherwise, run-aakomp will create one using either the provided --reference FASTA or a reference derived from the downloaded lineage.

Command-line Options

run-aakomp options:

Option	Description
--help-aakomp	Show help message for the aakomp binary and exit
--help-mibf	Show help message for the make_mibf binary and exit
-i, --input	Input genome file in FASTA format
-o, --output	Output prefix (default: _)
-r, --reference	Amino acid reference file (e.g., orthologous protein set)
-t, --threads	Number of threads to use (default: 48)
-v, --verbose	Enable verbose output
--debug	Enable debug mode for internal troubleshooting
-H, --hash	Number of hash functions used in miBF (default: 9)
-k, --kmer	Amino acid k-mer size (default: 9)
-l, --lower-bound	Minimum occupancy threshold for valid hits (default: 0.7)
--rescue-kmer	Number of consecutive k-mers to initiate a new seed (default: 4)
--max-offset	Maximum offset allowed when extending a seed during chaining (default: 2)
--lineage	Name of BUSCO lineage to auto-download and use as reference
--db-dir	Directory for or to store miBf database files (default: ./)
--dry-run	Print commands that would be executed, but don’t run them
--track-time	Record and report runtime statistics for each major step
--odb-version	BUSCO ortholog database version (default: 12)
--list-lineages	List all available BUSCO lineages and exit
--visualise	Visualise the cumulative distribution function
--version	Print version of aaKomp

License

aaKomp Copyright (c) 2025
British Columbia Cancer Agency Branch. All rights reserved.

Licensed under the GNU General Public License v3. See LICENSE or http://www.gnu.org/licenses/.

For commercial licensing inquiries, contact:
Patrick Rebstein – [email protected]