README.md

---
license: cc-by-nc-nd-4.0
task_categories:
- other
tags:
- chemistry
- mass-spectrometry
- DOM
- formula-assignment
- knn
- dissolved-organic-matter
- molecular-formula
pretty_name: DOM Formula Assignment Dataset
size_categories:
- n<1K
---

# DOM Formula Assignment Dataset

[![GitHub](https://img.shields.io/badge/GitHub-pcdslab/dom--formula--assignment--using--ml-blue?logo=github)](https://github.com/pcdslab/dom-formula-assignment-using-ml)

**Training and Testing Data for Machine Learning-Based Molecular Formula Assignment in Fulvic Acid DOM Mass Spectra**

> **Paper**: Under review

---

## Abstract

Dissolved organic matter (DOM) is a critical component of aquatic ecosystems, with the fulvic acid fraction (FA-DOM) exhibiting high mobility and ready bioavailability to microbial communities. While understanding the molecular composition is a vital area of study, the heterogeneity of the material, with a vast number of diverse compounds, makes this task challenging. Existing methods often struggle with incomplete formula assignment or reduced coverage highlighting the need for a better approach. In this study, we developed a machine learning approach using the k-nearest neighbors (KNN) algorithm to predict molecular formulas from ultra-high-resolution mass spectrometry data. The model was trained on chemical formulas assigned to multiple DOM samples using 7 Tesla(7T) and a 21 Tesla(21T) Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) system, and tested on an independent 9.4 T FT-ICR MS Fulvic Acid dataset. A synthetic dataset of plausible elemental combinations (C, H, O, N, S) was also generated to enhance generalization. Our approach achieved a 99.9% assignment rate on the labeled test set and assigned a total of 13,605 formulas for unlabeled peaks compared to the existing approach, which assigned 5914 formulas, achieving up to a 2.3X improvement in formula assignment coverage compared to existing methods.

---

## Dataset Description

This dataset contains training and testing data for Dissolved Organic Matter (DOM) formula assignment using K-Nearest Neighbors (KNN) machine learning models. The dataset includes mass spectrometry data from different sources and instruments (7T, 21T FT-ICR MS) used to train and evaluate KNN models for automated molecular formula assignment in DOM samples.

### Dataset Structure

```
├── DOM_training_set_ver2/          # 7T FT-ICR MS training data
│   ├── Table_Harney_River_*.xlsx   # Harney River samples (5 files)
│   ├── Table_Pantanal_*.xlsx       # Pantanal samples (2 files)
│   ├── Table_Suwannee_River_Fulvic_Acid_2.xlsx
│   └── Families_Tables/            # Chemical family classifications
│       ├── Families-Short-hr1.csv
│       ├── Families-Short-hr5.csv
│       ├── Families-Short-pantanal.csv
│       └── Families-Short-srfa.csv
│
├── DOM_training_set_ver3/          # 21T FT-ICR MS training data
│   ├── Table_Harney_River_*_21T.xlsx   # 21T Harney River samples (5 files)
│   └── Table_Suwannee_River_Fulvic_Acid_2_21T.xlsx
│
├── synthetic_data/                 # Synthetic formula combinations
│   └── formula_combinations_*.csv  # 51 files covering m/z ranges 100-610
│
├── DOM_testing_set/               # Testing samples
│   ├── Table_Pahokee_River_Fulvic_Acid.xlsx
│   ├── Table_Suwannee_River_Fulvic_Acid_2_v2.xlsx
│   └── Table_Suwannee_River_Fulvic_Acid_3.xlsx
│
├── DOM_testing_set_Peaklists/     # Detailed peaklists for testing
│   ├── PPFA/                       # Pahokee Peat Fulvic Acid
│   ├── SRFA2/                      # Suwannee River Fulvic Acid 2
│   └── SRFA3/                      # Suwannee River Fulvic Acid 3
```

---

## Data Preview

### Synthetic Data (formula_combinations_100-110.csv)
```csv
Formula,Mass_Daltons
CH9OS2,100.0023
CHO2N4,100.0027
C3H3O3N,100.0041
C2H3O2N3,100.0153
...
```

### Testing Set Peaklists (SRFA3_neg_8M_0.5s_5ppm_aFTk_PeakList_Rec.csv)
```csv
m/z Exp.,Intensity
101.0410446,0.002990723
101.045957,0.006607056
101.0480263,0.003494263
101.0494431,0.006072998
...
```

---

## Data Sources

### Training Data

- **7T FT-ICR MS Data (ver2)**: High-resolution mass spectrometry data from a 7 Tesla instrument
  - 8 Excel files with assigned formulas
  
- **21T FT-ICR MS Data (ver3)**: Ultra-high resolution data from a 21 Tesla instrument
  - 6 Excel files with assigned formulas
  - Higher mass accuracy and resolution

- **Synthetic Data**: Computationally generated formula combinations
  - 51 CSV files covering m/z range 100-610 Da
  - Used to expand training coverage for underrepresented m/z regions

### Testing Data

- **DOM Testing Set**: Independent samples for model validation
  - Pahokee Peat Fulvic Acid (PPFA)
  - Suwannee River Fulvic Acid standards (SRFA2, SRFA3)
  
- **Peaklists**: Detailed peak information with multiple acquisition methods
  - Different pulse sequences (aFTk, aFTsk, dmFTk)
  - CSV format with m/z and intensity data



---

## Usage

This dataset is designed to be used with the DOM KNN models available at:
[SaeedLab/dom-formula-assignment-using-knn](https://huggingface.co/SaeedLab/dom-formula-assignment-using-knn)

### Loading the Dataset

```python
from huggingface_hub import hf_hub_download
import pandas as pd

# Download a specific training file
file_path = hf_hub_download(
    repo_id="SaeedLab/dom-formula-assignment-data",
    filename="DOM_training_set_ver2/Table_Harney_River_1.xlsx",
    repo_type="dataset"
)

# Load with pandas
df = pd.read_excel(file_path)
print(df.head())
```

### Download Synthetic Data

```python
from huggingface_hub import hf_hub_download
import pandas as pd

# Download synthetic formula combinations
file_path = hf_hub_download(
    repo_id="SaeedLab/dom-formula-assignment-data",
    filename="synthetic_data/formula_combinations_200-210.csv",
    repo_type="dataset"
)

df = pd.read_csv(file_path)
print(f"Loaded {len(df)} synthetic formulas")
```

### Download Entire Dataset

```python
from huggingface_hub import snapshot_download

# Download all data
data_path = snapshot_download(
    repo_id="SaeedLab/dom-formula-assignment-data",
    repo_type="dataset"
)
print(f"Dataset downloaded to: {data_path}")
```

---

## Citation

This dataset supports research on automated formula assignment in DOM analysis using machine learning. A manuscript describing the methodology is currently under review.

---

## License

This model and associated code are released under the CC-BY-NC-ND 4.0 license and may only be used for non-commercial, academic research purposes with proper attribution. Any commercial use, sale, or other monetization of this model and its derivatives, which include models trained on outputs from the model or datasets created from the model, is prohibited and requires prior approval. Downloading the model requires prior registration on Hugging Face and agreeing to the terms of use. By downloading this model, you agree not to distribute, publish or reproduce a copy of the model. If another user within your organization wishes to use the model, they must register as an individual user and agree to comply with the terms of use. Users may not attempt to re-identify the deidentified data used to develop the underlying model. If you are a commercial entity, please contact the corresponding author.

---

## Contact

For any additional questions or comments, contact Fahad Saeed ([email protected]).

---

## Related Resources

- **Model Repository**: [SaeedLab/dom-formula-assignment-using-knn](https://huggingface.co/SaeedLab/dom-formula-assignment-using-knn)
- **GitHub Repository**: [pcdslab/dom-formula-assignment-using-ml](https://github.com/pcdslab/dom-formula-assignment-using-ml)