QSAR-Biodeg: Molecular Properties for Biodegradability Prediction

A subsampled dataset for quantitative structure-activity relationship (QSAR) modeling of chemical biodegradability. It was created by Eddie Bergman from the original qsar-biodeg dataset on OpenML using a controlled random sampling procedure. The subsampling parameters include a seed of 1, a maximum of 2000 rows, 100 columns, and 10 classes, with stratification applied.

Use Cases

• Predicting chemical biodegradability based on molecular property features.
• Training classification models to categorize chemicals into up to 10 biodegradability classes.
• Benchmarking feature selection algorithms on a dataset with 100 molecular descriptor columns.
• Developing QSAR models for environmental risk assessment of new compounds.

Strengths

• Subsampling was performed with a fixed random seed (1), ensuring reproducibility.
• The creation method used stratified sampling, which likely preserves class distribution.
• The dataset is derived from a known QSAR benchmark (qsar-biodeg) on OpenML.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

OpenML dataset qsar-biodeg (ID 1494), subsampled by Eddie Bergman.

Collection Method

Algorithmic subsampling with random column and row selection, preserving class stratification.