tags: protein corona nanoparticles stealth effect machine learning

Overview

PCSER is a computational tool for predicting protein corona stealth effects. It was built using the random forest machine learning approach.

📔 Documentation

Please check https://2003100127.github.io/pcser for how to use PCSER.

🛠️ Installation

PCSER can be installed in the following ways.

• (https://pypi.org/project/pcser)

  conda create --name pcser python=3.11
      
  conda activate pcser
  
  pip install pcser --upgrade
  

• 

  conda create --name pcser python=3.11
    
  conda activate pcser
  
  git clone https://github.com/2003100127/pcser.git
  
  cd pcser
  
  pip install .
  

🚀 Quick start

import pcser as pcs

pcs.load.evaluate(
    data_ref_fpn='./Proteomics_07262023_rv_C57BL6_spl54.xlsx',
    sv_fp='./',  # None to('data/')
    input_fpn='./example.xlsx',
    model_fpn='./best_cv.joblib',
    sheet_name='a', # a b
    # mfi_ref=[10271.33333, 10747, 10303.33333, 9663.333333, 10056],
    mfi_ref=[3606.333333, 3606.333333, 3606.333333, 3606.333333],

    # is_norm=True,
    # norm_met='minmax',  # minmax std maxabs
    # mode='compo',  # compo annot
    # mark='spl54',  # spl54 spl63
    # version='extended',  # extended old
)

Then, it outputs what is shown below.

# You are using extended sheets.
# You have selected the minmax normalization method.
# Data summary:
# Number of samples: 54
# Number of features: 419
# You have the samples: ['HuApoA1', 'MoApoA1', 'HuClusterin', 'MoClusterin']
# PCSER predictions: 
#              stealth_effect          MFI
# HuApoA1            0.670762  3099.790003
# MoApoA1            0.662108  3189.458730
# HuClusterin        0.634621  3474.270396
# MoClusterin        0.633914  3481.599008
# stealth_effect	MFI
# HuApoA1	0.670762	3099.790003
# MoApoA1	0.662108	3189.458730
# HuClusterin	0.634621	3474.270396
# MoClusterin	0.633914	3481.599008

📄 Citation

@article{PCSER,
    title = {PCSER},
    author = {Jianfeng Sun},
    doi = {xxx},
    url = {https://github.com/2003100127/pcser},
    journal = {xxx}
    year = {2024},
}

🏠 Homepage

📍Oxford University

📧 Reach us