---
pretty_name: TEA – Pathogen Identifier & Strain Tagger
tags:
  - biomedical
  - microbiology
  - pathogens
  - strains
  - full-text
  - named-entity-recognition
task_categories:
  - token-classification
language:
  - en
license:
  - mit
  - cc-by-4.0
  - cc-by-sa-4.0
  - cc-by-nc-4.0
  - cc-by-nc-sa-4.0
size_categories:
  - 1K<n<10K
dataset_info:
- config_name: pathogens
  features:
  - name: hash
    dtype: string
  - name: tokens
    sequence: string
  - name: ner_tags
    sequence:
      class_label:
        names:
          '0': O
          '1': B-COMM
          '2': I-COMM
          '3': B-PROB
          '4': I-PROB
          '5': B-OPPO
          '6': I-OPPO
          '7': B-NEGA
          '8': I-NEGA
          '9': B-STRA
          '10': I-STRA
          '11': B-PATH
          '12': I-PATH
  - name: locations
    dtype: string
  - name: license
    dtype: string
  - name: license_version
    dtype: string
  - name: license_url
    dtype: string
  - name: attribution_text
    dtype: string
  splits:
  - name: train
    num_bytes: 50995544
    num_examples: 505
  download_size: 11318610
  dataset_size: 50995544
- config_name: strains
  features:
  - name: hash
    dtype: string
  - name: tokens
    sequence: string
  - name: ner_tags
    sequence:
      class_label:
        names:
          '0': O
          '1': B-STRA
          '2': I-STRA
          '3': B-SPEC
          '4': I-SPEC
          '5': B-NEGA
          '6': I-NEGA
  - name: locations
    dtype: string
  - name: license
    dtype: string
  - name: license_version
    dtype: string
  - name: license_url
    dtype: string
  - name: attribution_text
    dtype: string
  splits:
  - name: train
    num_bytes: 41153821
    num_examples: 400
  download_size: 8836267
  dataset_size: 41153821
configs:
- config_name: pathogens
  description: Curated human microbial pathogen entities
  data_files:
  - split: train
    path: pathogens/train-*
  default: true
- config_name: strains
  description: Curated microbial strain and species entities
  data_files:
  - split: train
    path: strains/train-*
---

# Dataset Summary

**TEA – Pathogen Identifier & Strain Tagger** is a curated biomedical corpus with  
505 full-text articles curated for human microbial pathogens and 400 articles on microbial strains.  
Over 3,500 entity spans are annotated in IOB2 format.

**Usage**
```python
from datasets import load_dataset
pathogens_ds = load_dataset("tznurmin/tea_curated", "pathogens", split="train")
strains_ds = load_dataset("tznurmin/tea_curated", "strains", split="train")
print(pathogens_ds[152])
print(len(strains_ds))
```

## More information

  <https://github.com/tznurmin/TEA_curated_data>