CoMR: Reference Databases and Benchmarking for Mitochondrial Proteome Reconstruction

3.9 GB of reference databases, benchmarking resources, and a reproducibility container for the CoMR workflow. The deposit includes FASTA databases, BLAST resources, orthogroup alignments, HMM profiles, and benchmarking scripts for yeast and a protist. It was authored by Julie Boisard and last updated on 2026-04-21.

Use Cases

• Benchmarking mitochondrial protein prediction tools based on the provided ROC-AUC and precision-recall performance data.
• Reconstructing mitochondrial proteomes in divergent eukaryotic lineages using the integrative scoring framework described.
• Reproducing the comparative analysis between CoMR and TargetP2 for Saccharomyces cerevisiae and Paratrimastix pyriformis.
• Executing the CoMR Snakemake pipeline in a controlled environment using the provided Singularity/Apptainer container.

Strengths

• Includes a 3.9 GB collection of reference databases and benchmarking outputs.
• Benchmarking demonstrates a ROC-AUC of 0.92 in yeast and 0.86 in a divergent protist.
• Workflow is distributed as a containerized Snakemake pipeline for reproducible execution.
• Resources are provided under a permissive CC-BY-4.0 license.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the underlying data is unknown, which may limit suitability assessment.
• The description does not specify the taxonomic breadth of the reference databases beyond the two benchmark organisms.

Provenance

Source

figshare

Collection Method

Likely compiled and generated as part of the CoMR workflow development and benchmarking study.

Freshness

Last updated 2026-04-21 11:03:00; freshness should be verified.