# Cogeqc

> cogeqc

- Skill: `biomate-ai/cogeqc` (Agent Skill)
- Install (CLI): `npx skillmds@latest add biomate-ai/cogeqc`
- Raw SKILL.md: https://api.skillmd.com/api/skills/biomate-ai/cogeqc/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: biomate-ai (https://skillmd.com/u/biomate-ai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/biomate-ai/cogeqc

---


# cogeqc

## Workflows

### Standard Workflow

Assess genome assembly and annotation quality by comparing custom statistics against NCBI genomes and evaluating gene space completeness with BUSCO.

```r
library(cogeqc)

# 1. Retrieve and compare genome assembly statistics
maize_stats <- get_genome_stats(taxon = "Zea mays")
my_stats <- data.frame(
    accession = "my_lovely_maize",
    sequence_length = 2.4 * 1e9,
    gene_count_total = 50000,
    CC_ratio = 2
)
comparison <- compare_genome_stats(ncbi_stats = maize_stats, user_stats = my_stats)
plot_genome_stats(ncbi_stats = maize_stats, user_stats = my_stats)

# 2. Read and plot BUSCO summary statistics
output_dir <- system.file("extdata", package = "cogeqc")
busco_summary <- read_busco(output_dir)
plot_busco(busco_summary)
```

*Input: Custom genome assembly statistics and BUSCO output directories; Output: Quality control comparison tables and summary visualization plots.*

## When to Use
- Assessing genome assembly and annotation quality by comparing custom statistics against NCBI genomes using `get_genome_stats`, `compare_genome_stats`, and `plot_genome_stats`.
- Assessing gene space completeness using BUSCO via `run_busco`, `read_busco`, and `plot_busco`.
- Assessing orthogroup inference (e.g., from OrthoFinder or OrthoMCL) using a protein domain-based approach with `read_orthogroups` and `assess_orthogroups`.

## When NOT to Use
- For performing the actual genome assembly, gene annotation, or orthogroup clustering itself. Use external tools like Flye, Liftoff, or OrthoFinder, then use `cogeqc` for downstream quality control.

## Data Requirements
- For genome stats: A data frame of user-defined assembly statistics containing at least an `accession` column and other matching NCBI fields (e.g., `sequence_length`, `gene_count_total`, `CC_ratio`).
- For BUSCO: A directory containing BUSCO output files or pre-parsed BUSCO summary data frames.
- For orthogroups: An OrthoFinder `Orthogroups.tsv` file or a parsed data frame with columns `Orthogroup`, `Species`, and `Gene`.

## Key Parameters
- **taxon** (NULL): Taxon name or NCBI Taxonomy ID to retrieve statistics for in `get_genome_stats`.
- **filters** (NULL): List of key-value pairs to filter NCBI Datasets API results in `get_genome_stats`.
- **ncbi_stats** (NULL): Reference data frame of NCBI genome statistics in `compare_genome_stats` and `plot_genome_stats`.
- **user_stats** (NULL): Data frame of user-observed genome statistics to compare or highlight.
- **lineage** (NULL): Lineage dataset name for BUSCO in `run_busco`.
- **mode** (NULL): BUSCO run mode (e.g., "genome") in `run_busco`.
- **outpath** (NULL): Path to directory where BUSCO output will be stored in `run_busco`.
- **download_path** (NULL): Path to directory where BUSCO datasets will be downloaded in `run_busco`.

## Best Practices
- Use the `CC_ratio` (ratio of contigs to chromosome pairs) as a robust measurement of contiguity that allows cross-species comparisons.
- Filter NCBI reference genomes to match your assembly level (e.g., chromosome-scale) and annotation status using the `filters` parameter in `get_genome_stats`.
- Check if BUSCO is installed in your PATH using `busco_is_installed` before calling `run_busco`.
- Consider genomes with >90% complete BUSCOs as high quality when visualizing with `plot_busco`.

## Common Pitfalls
- `run_busco` failing due to missing system dependencies. Fix: Ensure BUSCO is installed and in your system PATH, or check with `busco_is_installed()`.
- Column mismatch when comparing custom stats with NCBI stats. Fix: Ensure the column names in your custom data frame exactly match the column names returned by `get_genome_stats` (e.g., `sequence_length` instead of `genome_size`).

## Alternatives
- `quast` (external tool) for general genome assembly quality metrics.
- `OrthoFinder` (external tool) for orthogroup inference.

## Citations
- Simão FA, Waterhouse RM, Ioannidis P, Kriventseva EV, Zdobnov EM (2015). BUSCO: Assessing Genome Assembly and Annotation Completeness with Single-Copy Orthologs. Bioinformatics.
- Wang P, Wang F (2022). A Proposed Metric Set for Evaluation of Genome Assembly Quality. Trends in Genetics.

## References
- Homepage: bioconductor.org/packages/cogeqc
- Vignette: https://bioconductor.org/packages/release/bioc/vignettes/cogeqc/inst/doc/genome_assembly_annotation_completeness.html