nadizik/3D_Synthetic_Petroleum_Derived_GEMS

3D Synthetic Petroleum-Derived GEMS (MVP Release)

📌 Dataset Overview

This dataset contains 197 elite, highly complex 3D molecular structures derived from petroleum fractions. Designed specifically for petrochemicals, materials science, organic semiconductors, and specialized additives, these compounds represent a curated "Golden Fund" of stable, complex hydrocarbons.

Unlike drug-like molecules, this dataset focuses on polycyclic architectures, rigid 3-ring systems, and dense carbon skeletons.

💡 Note on Dataset scale: This is an MVP release featuring our most complex, individually validated structures. If the community finds this data useful, a larger pool of 5,000 validated structures will be released. Let us know your thoughts in the "Community" tab!

🔬 How These Molecules Were Created & Validated

The generation pipeline followed a strict, multi-step workflow:

1. 1.Generative Modeling: Generation of novel paired/fused structures starting from specific hydrocarbon seeds (e.g., C1, C(N), N, O, Cc1).
2. 2.Individual 3D Validation: Each of these 197 structures underwent a dedicated, rigorous 3D conformation analysis and geometry optimization.
3. 3.Energy Minimization: Applied Universal Force Field (UFF) constraints to filter out unrealistic steric clashes or unstable shapes.
4. 4.Complexity Selection: Selected compounds with high synthetic complexity (SCI_Score averaging 6–7+) and unique structural properties.

📊 Features & Columns Description

The file smart_filtered_gems_3d_validated.csv contains the following columns (semicolon-separated):

• —SMILES: Flat chemical structure (SMILES string) of the hydrocarbon derivative.
• —Stability_Confidence: AI Critic confidence score regarding structural stability (all validated at 100%).
• —MolWt: Molecular weight of the compound (ranges from ~206 to ~320 g/mol, ideal for heavy hydrocarbon fractions).
• —LogP: Octanol-water partition coefficient, indicating high lipophilicity/hydrophobicity (ranges from 2.97 to 4.99).
• —Rings: Number of rings (focused strictly on rigid, stable $3$-ring architectures).
• —Seed_Used: The initial petrochemical seed/fragment used as a foundation by the generator.
• —UFF_Energy: Total potential energy after 3D conformation optimization (lower is more stable).
• —Energy_Per_Atom: Normalized UFF energy per individual atom, ensuring thermodynamic feasibility.

🛠️ How to Open the Dataset

🐍 Python (Pandas)

The dataset uses a semicolon (`;`) delimiter and dots (`.`) for decimal points.

import pandas as pd

# Correct way to load the dataset
df = pd.read_csv("smart_filtered_gems_3d_validated.csv", sep=";")
print(df.head())