README.md

---

license: cc-by-4.0
task_categories:
- fill-mask
- text-classification
- text2text-generation
- translation
- zero-shot-classification
tags:
- chemistry
- biology
- SMILES
size_categories:
- 1M<n<10M
pretty_name: 'Dosedo: DEL-DOS compounds'
configs:
- config_name: full
  data_files: "*/*.csv.gz"
  sep: '\t'
  default: true
- config_name: split
  sep: '\t'
  data_files:
  - split: train
    path: "train/*.csv.gz"
  - split: test
    path: "test/*.csv.gz"
  - split: validation
    path: "validation/*.csv.gz"
---

# hudson-2023-dosedo

SMILES of ~3.7 million diversity-oriented synthesis (DOS) compounds from a reported DNA-encoded library, in:

>

> Hudson, L., Mason, J.W., Westphal, M.V. et al. Diversity-oriented synthesis encoded by deoxyoligonucleotides. 
> Nat Commun 14, 4930 (2023). https://doi.org/10.1038/s41467-023-40575-5
>



The SMILES strings have been canonicalized, and split into training (70%), validation (15%), and test (15%) sets by Murcko scaffold. Additional features like molecular weight and topological polar surface area have also been calculated.

## Dataset Details

### Dataset Description

- **Curated by:** [@eachanjohnson](https://huggingface.co/eachanjohnson)
- **Funded by:** The Francis Crick Institute
- **License:** CC-by-4.0

### Dataset Sources

<!-- Provide the basic links for the dataset. -->

- **Repository:** https://doi.org/10.5281/zenodo.8136904
- **Paper** https://doi.org/10.1038/s41467-023-40575-5
<!-- - **Demo [optional]:** [More Information Needed] -->

## Uses

Developing chemistry models.

<!-- ### Direct Use   -->

<!-- This section describes suitable use cases for the dataset. -->

<!-- [More Information Needed]

### Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->

<!-- [More Information Needed]  -->

## Dataset Structure

- **SMILES**: SMILES string of compound
- **ROWNUM**: Row number of original dataset
- **id**: Numerical almost-unique identifier of compound
- **inchikey**: Unique identifier for compound
- **smiles**: RDKit-canonicalized SMILES string of compound
- **scaffold**: Murcko scaffold of compound
- **mwt**: Molecular weight of compound
- **clogp**: Crippen LogP of compound
- **tpsa**: Topological polar surface area of compound
- **is_train**: In training split

- **is_test**: In test split
- **is_validation**: In validation split



## Dataset Creation



### Curation Rationale



To make available a large dataset of SMILES strings for DOS compounds, as distinct from commonly encountered

virtual libraries from conventional combinatorial chemistry.



#### Data Collection and Processing



Data were processed using [schemist](https://github.com/scbirlab/schemist), a tool for processing chemical datasets.



#### Who are the source data producers?



Liam Hudson, Jeremy W. Mason, Matthias V. Westphal, Matthieu J. R. Richter, 

Jonathan R. Thielman, Bruce K. Hua, Christopher J. Gerry, Guoqin Xia, 

Heather L. Osswald, John M. Knapp, Zher Yin Tan, Praveen Kokkonda, 

Ben I. C. Tresco, Shuang Liu, Andrew G. Reidenbach, Katherine S. Lim, 

Jennifer Poirier, John Capece, Simone Bonazzi, Christian M. Gampe, Nichola J. Smith, 

James E. Bradner, Connor W. Coley, Paul A. Clemons, Bruno Melillo, C. Suk-Yee Hon, 

Johannes Ottl, Christoph E. Dumelin, Jonas V. Schaefer, Ann Marie E. Faust, 

Frédéric Berst, Stuart L. Schreiber, Frédéric J. Zécri & Karin Briner.



#### Personal and Sensitive Information



None.



<!-- ## Bias, Risks, and Limitations  -->



<!-- This section is meant to convey both technical and sociotechnical limitations. -->



<!--  [More Information Needed]  -->



<!-- ### Recommendations  -->



<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->



<!-- Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.  -->



## Citation



<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->



**BibTeX:**



```

@article{10.1038/s41467-023-40575-5, 

year = {2023}, 

title = {{Diversity-oriented synthesis encoded by deoxyoligonucleotides}}, 

author = {Hudson, Liam and Mason, Jeremy W. and Westphal, Matthias V. and Richter, Matthieu J. R. and Thielman, Jonathan R. and Hua, Bruce K. and Gerry, Christopher J. and Xia, Guoqin and Osswald, Heather L. and Knapp, John M. and Tan, Zher Yin and Kokkonda, Praveen and Tresco, Ben I. C. and Liu, Shuang and Reidenbach, Andrew G. and Lim, Katherine S. and Poirier, Jennifer and Capece, John and Bonazzi, Simone and Gampe, Christian M. and Smith, Nichola J. and Bradner, James E. and Coley, Connor W. and Clemons, Paul A. and Melillo, Bruno and Hon, C. Suk-Yee and Ottl, Johannes and Dumelin, Christoph E. and Schaefer, Jonas V. and Faust, Ann Marie E. and Berst, Frédéric and Schreiber, Stuart L. and Zécri, Frédéric J. and Briner, Karin}, 

journal = {Nature Communications}, 

doi = {10.1038/s41467-023-40575-5}, 

pmid = {37582753}, 

pmcid = {PMC10427684}, 

abstract = {{Diversity-oriented synthesis (DOS) is a powerful strategy to prepare molecules with underrepresented features in commercial screening collections, resulting in the elucidation of novel biological mechanisms. In parallel to the development of DOS, DNA-encoded libraries (DELs) have emerged as an effective, efficient screening strategy to identify protein binders. Despite recent advancements in this field, most DEL syntheses are limited by the presence of sensitive DNA-based constructs. Here, we describe the design, synthesis, and validation experiments performed for a 3.7 million-member DEL, generated using diverse skeleton architectures with varying exit vectors and derived from DOS, to achieve structural diversity beyond what is possible by varying appendages alone. We also show screening results for three diverse protein targets. We will make this DEL available to the academic scientific community to increase access to novel structural features and accelerate early-phase drug discovery. Most DNA-encoded library (DEL) syntheses are limited by the presence of sensitive DNA-based constructs. Here, the authors develop DOSEDO, a diverse 3.7 million compound DEL, generated through diversity-oriented synthesis that provides enhanced scaffold and exit vector diversity and gives validated binding hits for multiple protein targets.}}, 

pages = {4930}, 

number = {1}, 

volume = {14}

}

```



**APA:**



>

> Hudson, L., Mason, J. W., Westphal, M. V., Richter, M. J. R., Thielman, J. R., Hua, B. K., . . . Briner, K. (2023).

> Diversity-oriented synthesis encoded by deoxyoligonucleotides. Nature Communications, 14(1), 4930. 

> doi:10.1038/s41467-023-40575-5

>



<!-- ## Glossary [optional]  -->



<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->



<!-- [More Information Needed]



<!-- ## More Information [optional]



<!-- [More Information Needed]



<!-- ## Dataset Card Authors [optional]



<!-- [More Information Needed]  -->



## Dataset Card Contact



[@eachanjohnson](https://huggingface.co/eachanjohnson)