arunk95
commented
2 months ago Hello there, Thank you so much for reading our paper! The features that were used to build the random forest algorithm are included in the Human Proteome_30 features.csv file. That said, if you cannot find your protein in that list, it is probably because it did not fit the cut-offs or restrictions that were applied on certain feature calculations. For ex., proteins with <140 residues were not included in the PScore calculations (a restriction of the software itself). Features such as Disorder, or LCR% were directly calculated from the softwares provided their respective authors. If you'd like we could try to extract all features for your protein separately.
Hope this helps!
Dear Authors： thanks a lot for your wonderful job！ Now, I would have to predict my protein with your method. But it seems that the features are not contained in your features_list. It would be wonderful if you could provide your feature calculation script or detailed resources so that I can predict my protein stand-alone. Thank you very much!!