README.md

---
language: en
license: cc-by-nc-sa-4.0
pretty_name: IsItABarrel Database
tags:
- biology
- chemistry
dataset_summary: >-
  a curated database of approximately two million bacterial transmembrane beta
  barrels (TMBBs)
repo: https://isitabarrel.ku.edu/search
citation_bibtex: >-
  article{Montezano2023, title = {General features of transmembrane beta barrels
  from a large database}, volume = {120}, ISSN = {1091-6490}, url =
  {http://dx.doi.org/10.1073/pnas.2220762120}, DOI = {10.1073/pnas.2220762120},
  number = {29}, journal = {Proceedings of the National Academy of Sciences},
  publisher = {Proceedings of the National Academy of Sciences}, author =
  {Montezano,  Daniel and Bernstein,  Rebecca and Copeland,  Matthew M. and
  Slusky,  Joanna S. G.}, year = {2023}, month = jul}
citation_apa: >-
  Montezano, D., Bernstein, R., Copeland, M. M., & Slusky, J. S. G. (2023).
  General features of transmembrane beta barrels from a large database.
  Proceedings of the National Academy of Sciences of the United States of
  America, 120(29), e2220762120. doi:10.1073/pnas.2220762120
configs:
- config_name: TMBB_information
  data_files:
  - split: train
    path: TMBB_information/data/train-*
  - split: test
    path: TMBB_information/data/test-*
  - split: validate
    path: TMBB_information/data/validate-*
- config_name: fasta_files
  data_files:
  - split: train
    path: fasta_files/train.fasta
  - split: test
    path: fasta_files/test.fasta
  - split: validate
    path: fasta_files/val.fasta
- config_name: unsplit_TMBB_information
  data_files:
  - split: train
    path: unsplit_TMBB_information/data/train-*
dataset_info:
- config_name: TMBB_information
  features:
  - name: Protein_ID
    dtype: string
  - name: Organism
    dtype: string
  - name: Protein_Product
    dtype: string
  - name: Sequence_Length
    dtype: int64
  - name: SignalP5_Score
    dtype: float64
  - name: SignalP5_Type
    dtype: string
  - name: Representative
    dtype: string
  - name: Assigned
    dtype: int64
  - name: Protein_Type
    dtype: string
  - name: Strain
    dtype: string
  - name: Strand_Count
    dtype: float64
  - name: AlphaFoldDB_ID
    dtype: string
  - name: Sequence
    dtype: string
  splits:
  - name: train
    num_bytes: 891734109
    num_examples: 1163363
  - name: test
    num_bytes: 277747938
    num_examples: 387787
  - name: validate
    num_bytes: 277677434
    num_examples: 387785
  download_size: 1191909078
  dataset_size: 1447159481
- config_name: unsplit_TMBB_information
  features:
  - name: Protein_ID
    dtype: string
  - name: Organism
    dtype: string
  - name: Protein_Product
    dtype: string
  - name: Sequence_Length
    dtype: int64
  - name: SignalP5_Score
    dtype: float64
  - name: SignalP5_Type
    dtype: string
  - name: Representative
    dtype: string
  - name: Assigned
    dtype: int64
  - name: Protein_Type
    dtype: string
  - name: Strain
    dtype: string
  - name: Strand_Count
    dtype: float64
  - name: AlphaFoldDB_ID
    dtype: string
  - name: Sequence
    dtype: string
  splits:
  - name: train
    num_bytes: 1447159475
    num_examples: 1938935
  download_size: 1193205702
  dataset_size: 1447159475
dataset2_info:
  config_name: fasta_files
  features:
  - name: Protein_ID
    dtype: string
  - name: sequence
    dtype: string
  splits:
  - name: train
    num_bytes: 784000000
    num_examples: 2326726
  - name: test
    num_bytes: 242000000
    num_examples: 775574
  - name: validate
    num_bytes: 242000000
    num_examples: 775570
  download_size: 1268000000
  dataset_size: 1268000000
---
# IsItABarrel
This dataset contains 1,881,712 sequences collected from 600 different bacterial proteomes with sequences ranked by their likelihood of encoding a TMBB.

## QuickStart Usage

### Install HuggingFace Datasets package

Each subset can be loaded into python using the HuggingFace [datasets](https://huggingface.co/docs/datasets/index) library.  First, from the command line install the `datasets` library
 
    $ pip install datasets

Optionally set the cache directory, e.g.

    $ HF_HOME=${HOME}/.cache/huggingface/
    $ export HF_HOME

then, from within python load the datasets library
 
    >>> import datasets

### Load model datasets

To load one of the IsItABarrel model datasets, use `datasets.load_datase(...)`:


    >>> dataset = datasets.load_dataset("rmauder/IsItABarrel", "TMBB_information")

and the dataset is loaded as a `datasets.arrow_dataset.Dataset`

    >>> dataset
    <RESULT OF LOADING DATASET MODEL>

Which is a column oriented format that can be accessed directly, converted in to a `pandas.DataFrame`, or `parquet` format, e.g.

    >>> dataset.data.column("<colum_name>")
    >>> dataset.to_pandas()
    >>> dataset.to_parquet("dataset.parquet")

## Dataset Details
Dataset 1: `TMBB_information`
This dataset contains a csv file split into test/train/validate sets.  Each set contains 13 columns of information about each protein.

Dataset 2: `fasta_files`
This dataset contains pointers to fasta files that are already split into test/train/validate sets just like the csv files.  Each fasta file contains protein ID, organism, and the fasta sequence of the protein.

Dataset 3: `unsplit_TMBB_information`
This dataset contains a full unsplit csv file with information about every TMBB protein.  This can be used if anyone would like to test different splits or to filter data further based on organism or TMBB type.

### Dataset Description 
This dataset contains information like protein_ID and protein type for almost two million transmembrane beta barrel proteins.  It contains both known proteins and predicted proteins based on prokaryotic genomes.

- **Acknowledgements:** Daniel Montezano, Rebecca Bernstein, Matthew M. Copeland, and Joanna S. G. Slusky

- **License:** cc-by-nc-sa 4.0

### Dataset Sources
- **Repository:** [https://isitabarrel.ku.edu/search](https://isitabarrel.ku.edu/search)
- **Paper:**
  
    Montezano, D., Bernstein, R., Copeland, M. M., & Slusky, J. S. G. (2023).
    General features of transmembrane beta barrels from a large database. Proceedings
    of the National Academy of Sciences of the United States of America, 120(29), e2220762120.
    doi:10.1073/pnas.2220762120


### Source Data 
The source data is set up as a large downloadable file that is not clustered or split into train test validate splits.

## Citation 
@article{Montezano2023, title = {General features of transmembrane
  beta barrels from a large database}, volume = {120}, ISSN = {1091-6490}, url = {http://dx.doi.org/10.1073/pnas.2220762120},
  DOI = {10.1073/pnas.2220762120}, number = {29}, journal = {Proceedings of the National
  Academy of Sciences}, publisher = {Proceedings of the National Academy of Sciences},
  author = {Montezano,  Daniel and Bernstein,  Rebecca and Copeland,  Matthew M. and
  Slusky,  Joanna S. G.}, year = {2023}, month = jul}

## Dataset Card Authors
Ryan Mauder rmauder@butler.edu