DeepNGlyPred: a Deep Neural Network-based approach for human N-linked Glycosylation site prediction Please cite
Pakhrin, S.C.; Aoki-Kinoshita, K.F.; Caragea, D.; KC, D.B. DeepNGlyPred: A Deep Neural Network-Based Approach for Human N-Linked Glycosylation Site Prediction. Molecules 2021, 26, 7314. https://doi.org/10.3390/molecules26237314
Programs were executed using anaconda version: 2020.07, recommended to install the same
The programs were developed in the following environment. python : 3.8.3.final.0, python-bits : 64, OS : Linux, OS-release : 5.8.0-38-generic, machine : x86_64, processor : x86_64, pandas : 1.0.5, numpy : 1.18.5, pip : 20.1.1, scipy : 1.4.1, scikit-learn : 0.23.1., keras : 2.4.3, tensorflow : 2.3.1.
-
Please place the DNN_trained_with_nGlyde_test_with_nglyde.ipynb, 0.605 ANN molecules.ipynb, 447-feature_test.xlsx, All_feature_Independent_dataframe.csv, trained models like 3080_NetSurfP2.0_PSSM_GD8.h5 and Glycosylation_model__26__.h5 in the same directory where you will execute the python program.
-
Run DNN_trained_with_nGlyde_test_with_nglyde.ipynb, 0.605 ANN molecules.ipyn to obtain reported results.
*** For your convenience a sample PSSM file (P48066.PSSM) for protein P48066 generated using PSI-BLAST, a sample NetsurfP-2.0 file for protein B7Z4J0, B7Z4J0.fasta.csv generated by NetSurfP-2.0, and Ratio_Normalized_NGlyDE.txt file which is generated by Gapped Dipeptide feature extraction program has been uploaded ***
If you need any futher help please contact Dr. Dukka B. KC at [email protected].