https://github.com/animesh/deepcollisionalcrosssection

https://github.com/animesh/deepcollisionalcrosssection

Repository

https://github.com/animesh/DeepCollisionalCrossSection/blob/runpy37tf13/

## Setup 
```
wget https://repo.anaconda.com/archive/Anaconda3-2021.05-Linux-x86_64.sh
sh Anaconda3-2021.05-Linux-x86_64.sh
conda create -n py37 python=3.7 anaconda
conda activate py37
sudo apt install libfreetype-dev
sudo apt-get install libfontconfig1-dev
sudo apt-get install libopenblas-dev
sudo apt-get install libhdf5-dev
sudo apt install python3-pip
pip install -r requirements.txt
pip install tensorflow-gpu==1.13.2
pip install  twisted
conda -c rapidsai -c nvidia -c conda-forge -c defaults rapids-blazing=0.17 cudatoolkit=10.1 cudatoolkit=11.0 cudatoolkit=10.0 cudnn
git clone https://github.com/animesh/DeepCollisionalCrossSection
git checkout 3cec81c7992536200844f0f6527a076e662ff842
cd DeepCollisionalCrossSection
sudo apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub
sudo sh -c 'echo "deb http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64 /" > /etc/apt/sources.list.d/cuda.list'
sudo apt update
sudo apt upgrade
sudo apt-get install -y cuda-toolkit-10-0
#sudo apt-get install -y cuda-toolkit-11-0
#sudo apt install nvidia-utils-460
#sudo apt-get install cuda-drivers
export CUDA_VISIBLE_DEVICES=0
```

## Create dataset with MaxQuant Score instead of CCS
```
python process_data_final.py evidence.txt
```

## Profit!
```
mkdir out
python bidirectional_lstm.py evidence.txt_proc_2_train.pkl evidence.txt_proc_2_test.pkl
```

Output will be written to subfolder "out" 



# CCS Model Training and Prediction

Publication:
- doi: https://doi.org/10.1101/2020.05.19.102285
- biorxiv: https://www.biorxiv.org/content/10.1101/2020.05.19.102285v1

## Library Setup

Setup CUDA 10.0 with cudnn and install the required python libraries with pip:

```
pip install -r requirements.txt
```

## Prediction with Pre-Trained Model

Unzip the checkpoint found in out.

Prepare a csv file that contains Sequence and Charge Information and use the provided `predict.py` script:
```
python predict.py  
```
For the format see the provided example file in `./data/combined_reduced.csv`

## Process data
Use the provided notebook: `process_data_final.ipynb`

It uses the raw data files and saves train and test files in pkl format to disc in `./data_final`

## Training

The `bidirectional_lstm.py` file contains training and prediction routines.

Training is done by setting the paths in `run_training.py` and executing it.
The complete dataset will be uploaded at a later stage of publication.

## Evaluation

Use the provided `evaluate.ipynb` Jupyter Notebook.

Owner

GitHub Events