MelonnPan, MiMeNet, SparseNED: Metagenomics-to-Metabolomics Model Performance Comparison

Bianca-Maria Cosma's dataset compares data processing methods for three metagenomics-to-metabolomics models: MelonnPan, MiMeNet, and SparseNED. The 9.7 KB XLSX file contains performance results measured using the average Spearman's rank correlation coefficient for the 50 best predicted features. It was last updated on April 10, 2026.

Use Cases

• Benchmarking the performance of the MelonnPan, MiMeNet, and SparseNED models based on the reported Spearman's correlation coefficients.
• Evaluating the impact of different data processing approaches, such as arcsine and quantile transformations, on model predictions.
• Identifying the optimal data processing pipeline for a given metagenomics-to-metabolomics prediction task based on the comparative results.

Strengths

• Performance is measured using a standardized metric: the average Spearman's rank correlation coefficient for the 50 best predicted features.
• Explicitly compares the 'default' internal data processing approach of each model against other methods.
• Dataset is small (9.7 KB) and in a widely accessible XLSX format, facilitating quick inspection.

Limitations

• Row count and column-level documentation are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality and structure require manual inspection after download.
• Some experiments were omitted, such as not repeating transformations already part of a model's default pipeline.

Provenance

Source

Bianca-Maria Cosma via figshare.

Collection Method

Likely contains results from computational experiments comparing model performance under different data processing conditions.

Time Range

null

Freshness

Last updated 2026-04-10 17:48:57; freshness should be verified.

Geography

null