File size: 1,774 Bytes
e2985fa bea5c46 0667e0f bea5c46 e8450a5 cd0b4bb e83eab3 e2985fa cd0b4bb 57a1087 e8450a5 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 |
The CHOP_inhibitors_SMILES_1H_NMR_spectrosopy_concise dataset is a part of the study "Predicting Nanoparticle Effects on Small Biomolecule Functionalities Using the Capability of Scikit-learn and PyTorch: A Case Study on Inhibitors of the DNA Damage-Inducible Transcript 3 (CHOP)"
https://doi.org/10.48550/arXiv.2504.09537
Dataset content: The Dataset has 20,309 rows of samples, 12 columns of features. Each feature corresponds to a range of the 1H NMR spectroscopy chemical shifts scale. Natural numbers have defined these ranges. The values of the features are the number of chemical picks in the corresponding range/feature for the relevant compound. The column "target" contains the labels, as follows: 1 for the active CHOP inhibitors and 0 for the inactive inhibitors.
Dataset generation: The labels and the corresponding SMILES were taken from PubChem AID 2732 bioassay https://pubchem.ncbi.nlm.nih.gov/bioassay/2732 "HTS for small molecule inhibitors of CHOP to regulate the unfolded protein response to ER stress", which contains 8,241 active and 210,429 inactive compounds.
The inactive samples were reduced to 12,091 samples by merging the aforementioned dataset with the dataset of PubChem AID 1996 bioassay https://pubchem.ncbi.nlm.nih.gov/bioassay/1996 "Aqueous Solubility from MLSMR Stock Solutions", on SMLES, keeping only the common compounds for both bioassays.
The SMILES notations on the remaining samples were then converted into 1H NMR spectroscopy data using the NMRDB online tool https://www.nmrdb.org/new_predictor/index.shtml?v=v2.171.1, counting the 1H NMR spectroscopy chemical shifts in ranges defined by natural numbers.
---
license: apache-2.0
authors:
- Mariya L. Ivanova
- Nicola Russo
- Konstantin Nikolic
- Gueorgui Mihaylov
--- |