https://github.com/singjc/redeem

Repository for Deep Learning Models for Mass Spectrometry written in Rust
https://github.com/singjc/redeem

deep-learning mass-spectrometry rust semi-supervised-learning

Last synced: 7 months ago
JSON representation

Repository for Deep Learning Models for Mass Spectrometry written in Rust

• Host: GitHub
• URL: https://github.com/singjc/redeem
• Owner: singjc
• Created: 2024-10-09T18:13:53.000Z (about 1 year ago)
• Default Branch: master
• Last Pushed: 2025-03-13T19:27:01.000Z (7 months ago)
• Last Synced: 2025-03-13T19:31:42.320Z (7 months ago)
• Topics: deep-learning, mass-spectrometry, rust, semi-supervised-learning
• Language: Rust
• Homepage:
• Size: 326 KB
• Stars: 11
• Watchers: 4
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          


  

    

    

    

  



---

# ReDeeM: Repository for Deep Learning Models for Mass Spectrometry

ReDeeM is a Rust crate designed for implementing deep learning models specifically tailored for mass spectrometry data. The primary goal of this project is to facilitate the prediction of peptide properties and to develop classifier scoring models (TDA). 

### Usage

The ReDeeM crates are designed to be used as a library in other projects, i.e. in Sage. To use the ReDeeM crates, add the following to your `Cargo.toml` file:

```toml

[dependencies]

redeem-properties = { git = "https://github.com/singjc/redeem.git", branch = "master" }

redeem-classifiers = { git = "https://github.com/singjc/redeem.git", branch = "master" }

```

**Note**: The ReDeeM crates are still under development and are not yet available on crates.io.

### Current Crates

The ReDeeM project consists of two primary crates:

1. **redeem-properties**: 

   - This crate focuses on deep learning models for peptide property prediction. It implements models for predicting retention time (RT), ion mobility (IM), and MS2 fragment intensities using the Candle library.

   - The models can be fine-tuned on new data and can be saved in the safetensor format for later use.

   

   - Current Models

  

    Model | Name | Architecture | Implemented

    --- | --- | --- | ---

    AlphaPept RT Model | `redeem_properties::RTCNNLSTMModel` | CNN-LSTM | :heavy_check_mark:

    AlphaPept MS2 Model | `redeem_properties::MS2BertModel` | Bert | :heavy_check_mark:

    AlphaPept IM Model | `redeem_properties::CCSCNNLSTMModel` | CNN-LSTM | :heavy_check_mark:

2. **redeem-classifiers**:

   - This crate is aimed at developing semi-supervised scoring classifier models. The goal is to create models for separating target peptides from decoys.

  

   - Current Models

  

    Model | Name | Architecture | Implemented

    --- | --- | --- | ---

    XGBoost Classifier | `redeem_classifiers::XGBoostClassifier` | XGBoost | :heavy_check_mark:

    GBDT Classifier | `redeem_classifiers::GBDTClassifier` | GBDT | :heavy_check_mark:

    SVM Classifier | `redeem_classifiers::SVMClassifier` | SVM | :heavy_check_mark:

> [!NOTE]

> To use the XGBoost classifier, or the SVM classifier, you need to compile with the `--features xgboost` or `--features linfa` flag respectively.

> [!IMPORTANT]

> The XGBoost crate is a wrapper around the original XGBoost library, which requires clang/c++ to be installed on the system. On Ubuntu, you can do the following:

    

    ```bash

    sudo apt update

    sudo apt install build-essential

    sudo apt install clang

    sudo apt install libstdc++-12-dev

    ```