gbyuvd/coconut-chembl34-mol-sim

Dataset Card for COCONUT+ChemBL34 SELFIES for Sentence Similarity Training

This dataset is a collection of generated molecular structures pairs represented as SELFIES (Self-Referencing Embedded Strings) with MACCS fingerprint cosine similarity as label, created by combining and processing data from COCONUTDB and ChemBL34. It contains 7M pairs in total with ~5M trainable across 6 chunks.

The dataset is specifically designed for fine-tuning a sentence transformer for similarity task on the pretrained ChemFIE base model.

Key features:

• Source: combined data from COCONUTDB (natural compounds) and ChemBL34 (bioactive compounds)
• Format: score, text1, text2
• Size: 7,176,372 pairs
• Purpose: sentence transformer training/fine-tuning for similarity task
• Similarity Label: cosine similarity of MACCS fingerprints generated from RDKit

Disclaimer: For Academic Purposes Only

The information and dataset provided is for academic purposes only. It is intended for educational and research use, and should not be used for any commercial or legal purposes. The author do not guarantee the accuracy, completeness, or reliability of the information.

Dataset Details

Out-of-Scope Use

Any purposes that could lead to the creation or promotion of harmful substances.

Dataset

Structure

Total

Dataset	Reference	Total Number of Pairs
COCONUTDB	(Sorokina et al. 2021)	1,183,186
ChemBL34	(Zdrazil et al. 2023)	5,993,186
Total		7,176,372

Training

Dataset	File Name	Number of Pairs
COCONUTDB	NP_train_data	1,049,286
Combined (Part I)	atrain_data_chunk1	1,100,000
Combined (Part II)	atrain_data_chunk2	1,100,000
Combined (Part III)	atrain_data_chunk3	1,100,000
Combined (Part IV)	atrain_data_chunk4	1,100,000
Combined (Part V)	atrain_data_chunk5	394,548
Total		5,843,834

Test

Dataset	File Name	Number of Pairs
COCONUTDB	NP_test_data	299,659
Combined Test (comb from test splits of COCONUTDB & ChemBL34)	atest_data	898,979

Source Data

The dataset combines two sources of molecular data before sentence pairs generation:

1. Natural compounds from COCONUTDB (Sorokina et al., 2021)
2. Bioactive compounds from ChemBL34 (Zdrazil et al., 2023)

Data Collection and Processing

1. Data Collection: Raw data was gathered from COCONUTDB and ChemBL34.

2. Data Preprocessing:

   • Internal duplicates were removed.
   • Molecules were converted to SELFIES representation.
   • Molecules were filtered to ensure token length ≤ 510 (512 - 2 special tokens)
   • SELFIES were converted back to SMILES.

3. Data Merging:

   • The processed datasets were merged into a combined dataset.
   • Duplicates were removed again.

Who are the source data producers?

The source data producers are the contributors to COCONUTDB and ChemBL34 databases.

Annotations

The dataset was prepared using various methods to create informative molecule pairs and labels for training a SELFIES-based sentence transformer.

Annotation process

4. Tokenization and Encoding (for binning):

   • Using a custom tokenizer that was trained on the SELFIES representations, SELFIES were encoded into numerical vectors.

5. Complexity Score Calculation:

   • A ratio was calculated for each molecule: sum of encoded IDs / length of SELFIES.
   • This ratio serves as a simple complexity score (similar to what was done in the base model)

6. Binning:

   • The complexity scores were log-transformed to reduce skewness.
   • Sturges' formula was used to determine the optimal number of bins.
   • Molecules were assigned to bins based on their log-transformed complexity scores.

7. Pair Generation:

   • A target size for each bin was calculated based on the average bin size.
   • For in-strata pairs:
     • Molecules from the same bin were paired.
     • Bins larger than the target size were undersampled.
     • Bins smaller than the target size were oversampled.
   • For inter-strata pairs:
     • Pairs were created between bins with an absolute difference > 3.
   • The pairing process aimed to balance the distribution across complexity levels.
   • Empty subsets were handled to ensure robust processing.

8. Similarity Labeling:

   • MACCS fingerprints were generated for each molecule using RDKit.
   • Cosine similarity between the MACCS fingerprints was calculated using CuPy for GPU acceleration.
   • Error handling was implemented for cases where molecules couldn't be processed.
   • The labels were resampled to ensure a balanced range of examples ([0,1])
   • The similarity scores were rounded to two decimal places for training.

9. Splitting:

   • The dataset was split into 80% for training, 10% for validation, and 10% for testing.
   • For the Natural Products set (NP), the test set is labeled "NP-iso-base" because the SMILES strings were not canonicalized (isomeric forms were retained). _ file name: NP_test_data.csv
   • The test set from ChemBL34 was then combined with the NP test set, and this combined set is referred to as "combined." _ file name: atest_data.csv

This methodology aims to provide a diverse set of molecule pairs with labels indicating structural similarity. The combination of complexity binning, balanced inter- and intra-strata sampling, and MACCS fingerprint similarity labeling is intended to capture a range of molecular complexities while providing chemically relevant labels for model training.

Bias, Risks, and Limitations

• The dataset may be biased towards certain types of molecules based on the focus of COCONUTDB and ChemBL34.
• The dataset does not include filtering by Lipinski's Rule of Five, which may include molecules that are not typically considered drug-like.
• The pairing and annotation methods may lead to certain bias and imbalance, the data distribution should be examined before use.
• The chunking process may introduce some artifacts in the data distribution.

Recommendations

Users should be aware of the potential biases in the dataset and consider the implications for their specific use case.

Citation

If you find this project useful in your research and wish to cite it, please use the following BibTex entry:

ChemFIE Project:

@software{chemfie_basebertmlm,
  author = {GP Bayu},
  title = {{ChemFIE Base}: Pretraining A Lightweight BERT-like model on Molecular SELFIES},
  url = {https://huggingface.co/gbyuvd/chemselfies-base-bertmlm},
  version = {1.0},
  year = {2024},
}

ChEMBL34:

@misc{chembl34,
  title={ChemBL34},
  year={2023},
  doi={10.6019/CHEMBL.database.34}
}

@article{zdrazil2023chembl,
  title={The ChEMBL Database in 2023: a drug discovery platform spanning multiple bioactivity data types and time periods},
  author={Zdrazil, Barbara and Felix, Eloy and Hunter, Fiona and Manners, Emma J and Blackshaw, James and Corbett, Sybilla and de Veij, Marleen and Ioannidis, Harris and Lopez, David Mendez and Mosquera, Juan F and Magarinos, Maria Paula and Bosc, Nicolas and Arcila, Ricardo and Kizil{\"o}ren, Tevfik and Gaulton, Anna and Bento, A Patr{\'i}cia and Adasme, Melissa F and Monecke, Peter and Landrum, Gregory A and Leach, Andrew R},
  journal={Nucleic Acids Research},
  year={2023},
  volume={gkad1004},
  doi={10.1093/nar/gkad1004}
}

COCONUTDB:

@article{sorokina2021coconut,
  title={COCONUT online: Collection of Open Natural Products database},
  author={Sorokina, Maria and Merseburger, Peter and Rajan, Kohulan and Yirik, Mehmet Aziz and Steinbeck, Christoph},
  journal={Journal of Cheminformatics},
  volume={13},
  number={1},
  pages={2},
  year={2021},
  doi={10.1186/s13321-020-00478-9}
}

For the SELFIES representation used in data processing:

@article{krenn2020selfies,
  title={Self-referencing embedded strings (SELFIES): A 100\% robust molecular string representation},
  author={Krenn, Mario and H{\"a}se, Florian and Nigam, AkshatKumar and Friederich, Pascal and Aspuru-Guzik, Alan},
  journal={Machine Learning: Science and Technology},
  volume={1},
  number={4},
  pages={045024},
  year={2020},
  doi={10.1088/2632-2153/aba947}
}

Dataset Card Contact

G Bayu (gbyuvd@proton.me)

This project has been quiet a journey for me, I’ve dedicated hours on this and I would like to improve myself, this data, and future projects. However, financial and computational constraints can be challenging.

If you find my work valuable and would like to support my journey, please consider supporting me here. Your support will help me cover costs for computational resources, data acquisition, and further development of this project. Any amount, big or small, is greatly appreciated and will enable me to continue learning and explore more.

Thank you for checking out this project, I am more than happy to receive any feedback, so that I can improve myself and the future projects I will be working on.