PlantMarkerBench: Multi-Species Literature Evidence for Plant Genes and Cell Types

PlantMarkerBench is a sentence-level benchmark for attributing evidence about plant markers from biological literature. It contains 5,550 evidence instances covering 1,036 unique genes and 127 observed cell types across Arabidopsis, maize, rice, and tomato. The dataset was created by Sajib-006 and last updated on Hugging Face in May 2026.

Use Cases

• Train named entity recognition models to identify plant genes and cell types based on sentence-level evidence.
• Benchmark evidence attribution models for plant biology based on literature-grounded instances.
• Develop multi-species text mining tools for Arabidopsis, maize, rice, and tomato based on the described coverage.
• Analyze the distribution of marker evidence across different plant species based on the benchmark structure.

Strengths

• Contains 5,550 sentence-level evidence instances, providing a substantial corpus for model training.
• Covers 1,036 unique genes and 127 observed cell types, offering a multi-faceted biological scope.
• Spans four distinct plant species (Arabidopsis, maize, rice, tomato), enabling cross-species analysis.

Limitations

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

Provenance

Source

Hugging Face dataset by author Sajib-006.

Collection Method

Likely extracted from full-text biological literature.

Time Range

null

Freshness

Last updated 2026-05-11 05:25:48; freshness should be verified.

Geography

null