Collabrated with Peking University People's Hospital, we built diagnostic model for Rheumatoid Arthritis (RA) based on protein array. With 44 selected autoantibodies as markers, the model has achieved high accuracy (AUC=0.86, Specificity=0.9, Sensitivity=0.68). Here, we provided the model and script for RA prediction.
- python3
- numpy
- sklearn
- pickle
- pandas
git clone https://github.com/gao-lab/RA_prediction.git
python RA_prediction.py example.txt output.txt
*Both input file and output file are REQUIRED.
See example.txt. The first row is the patient IDs and should be started with "#". The first column is the IDs of the autoantibody markers (RFE.select.ids). Note that the autoantibody expression value in protein array shoulde be with the same rank with RFE.select.ids.
Yu-Jian Kang <[email protected]>