ATPs / 3D-MSNet

A point cloud based deep learning model for untargeted feature detection and quantification in profile LC-HRMS data

3D-MSNet: A point cloud based deep learning model for untargeted feature detection and quantification in profile LC-HRMS data

Highlights

• Novelty: 3D-MSNet enables direct spatial analysis on lossless 3D MS data for the first time, considering the feature extraction problem as an instance segmentation task on LC-MS point clouds.
• Accuracy: 3D-MSNet achieved the best performance in feature detection and quantification compared to popular software in metabolomics and proteomics.
• Reliability: 3D-MSNet achieved the best performance on all the three benchmark datasets (metabolomics TripleTOF 6600, metabolomics QE HF, proteomics Orbitrap XL) with the same pre-trained model (trained on metabolomics TripleTOF 6600).
• Efficiency: 3D-MSNet spent similar analysis time as traditional methods and about five times faster than other deep-learning-based methods.
• Open source: We open-sourced 3D-MSNet in order to promote the accuracy of MS data interpretation more broadly.
• Dataset: We provide open access to our training dataset, named the 3DMS dataset. Each signal point in the 3DMS dataset was manually annotated with an instance label, indicating whether the point belongs to a feature and to which feature it belongs.

Sample video

3D-MSNet_sample_video_480.mov

Environment

Recommended

Intel(R)_Core(TM)_i9-10900K CPU, 32GB memory, GeForce RTX 3090 GPU

Ubuntu 16.04 + CUDA 11.1 + cuDNN 8.0.5

Anaconda 4.9.2 + Python 3.6.13 + PyTorch 1.9

Setup

1. Prepare the deep-learning environment based on your system and hardware, including GPU driver, CUDA, cuDNN, Anaconda, Python, and PyTorch.

2. Install the dependencies. Here we use ROOT_PATH to represent the root path of 3D-MSNet.

   cd ROOT_PATH

   pip install -r requirements.txt

3. Compile CUDA code. This will take a few minutes.

   cd cuda

   python setup.py install

Datasets

The 3DMS dataset and all the benchmark datasets (mzML format) can be freely downloaded at Zenodo.

Raw MS files of the metabolomics datasets can be downloaded at Google Drive.

Raw MS files of the proteomics datasets can be downloaded at ProteomeXchange (dataset PXD001091).

Targeted annotation results, evaluation results and evaluation methods can be downloaded at Zenodo.

Run 3D-MSNet

Demos

Our demos can help you reproduce the evaluation results.

Place the benchmark datasets as follows.

3D-MSNet-master
├── dataset
│   ├── TripleTOF_6600
│   │   ├── mzml
│   │   │   ├── *.mzML

│   ├── QE_HF
│   │   ├── mzml
│   │   │   ├── *.mzML

│   ├── Orbitrap_XL
│   │   ├── mzml
│   │   │   ├── *.mzML

Then run scripts in folder DEMO. For example:

cd ROOT_PATH

Prepare point clouds: python DEMO/TripleTOF_6600_untarget/0_pc_extraction.py

Extract features: python DEMO/TripleTOF_6600_untarget/1_peak_detection.py

The result files are saved in the dataset folder.

Customized running

Refer to DEMO for parameter setting of different LC-MS platforms.

cd ROOT_PATH

Prepare point clouds:

python workflow/predict/point_cloud_extractor.py --data_dir=PATH_TO_MZML --output_dir=POINT_CLOUD_PATH --window_mz_width=0.8 --window_rt_width=6 --min_intensity=128 --from_mz=0 --to_mz=2000 --from_rt=0 --to_rt=300 --expansion_mz_width=0.1 --expansion_rt_width=1

Extract features:

python workflow/predict/main_eval.py --data_dir=POINT_CLOUD_PATH --mass_analyzer=orbitrap --mz_resolution=60000 --resolution_mz=400 --rt_fwhm=0.1 --target_id=None

Run python workflow/predict/point_cloud_extractor.py -h and python workflow/predict/main_eval.py -h to learn parameter details.

Train

We provided a pretrained model in experiment folder.

If you want to train the model on your self-annotated data, prepare your .csv files refer to the 3DMS dataset. Each MS signal should be annotated an instance label.

Place the training dataset as follows.

3D-MSNet-master
├── dataset
│   ├── your_training_dataset
│   │   ├── dataset_anno
│   │   │   ├── [id_mz_rt].csv

Then change the training parameters at config/msnet_default.yaml

cd ROOT_PATH

Split training set and validation set:

python workflow/train/dataset_generator.py

Start training:

python workflow/train/main_train.py

Trained models are saved in experiment folder.

Citation

Cite our paper at:

@article{10.1093/bioinformatics/btad195,
    author = {Wang, Ruimin and Lu, Miaoshan and An, Shaowei and Wang, Jinyin and Yu, Changbin},
    title = "{3D-MSNet: a point cloud-based deep learning model for untargeted feature detection and quantification in profile LC-HRMS data}",
    journal = {Bioinformatics},
    volume = {39},
    number = {5},
    year = {2023},
    month = {04},
    abstract = "{Liquid chromatography coupled with high-resolution mass spectrometry is widely used in composition profiling in untargeted metabolomics research. While retaining complete sample information, mass spectrometry (MS) data naturally have the characteristics of high dimensionality, high complexity, and huge data volume. In mainstream quantification methods, none of the existing methods can perform direct 3D analysis on lossless profile MS signals. All software simplify calculations by dimensionality reduction or lossy grid transformation, ignoring the full 3D signal distribution of MS data and resulting in inaccurate feature detection and quantification.On the basis that the neural network is effective for high-dimensional data analysis and can discover implicit features from large amounts of complex data, in this work, we propose 3D-MSNet, a novel deep learning-based model for untargeted feature extraction. 3D-MSNet performs direct feature detection on 3D MS point clouds as an instance segmentation task. After training on a self-annotated 3D feature dataset, we compared our model with nine popular software (MS-DIAL, MZmine 2, XCMS Online, MarkerView, Compound Discoverer, MaxQuant, Dinosaur, DeepIso, PointIso) on two metabolomics and one proteomics public benchmark datasets. Our 3D-MSNet model outperformed other software with significant improvement in feature detection and quantification accuracy on all evaluation datasets. Furthermore, 3D-MSNet has high feature extraction robustness and can be widely applied to profile MS data acquired with various high-resolution mass spectrometers with various resolutions.3D-MSNet is an open-source model and is freely available at https://github.com/CSi-Studio/3D-MSNet under a permissive license. Benchmark datasets, training dataset, evaluation methods, and results are available at https://doi.org/10.5281/zenodo.6582912.}",
    issn = {1367-4811},
    doi = {10.1093/bioinformatics/btad195},
    url = {https://doi.org/10.1093/bioinformatics/btad195},
    note = {btad195},
    eprint = {https://academic.oup.com/bioinformatics/article-pdf/39/5/btad195/50305059/btad195.pdf},
}

License

3D-MSNet is an open-source tool, using Mulan Permissive Software License，Version 2 (Mulan PSL v2)