Ensemble Methods

1. The program will automatically grab any .csv files in the directory so only all necessary files for a given run should be in there e.g., ABCpred outputs for EBV training, Bepipred2 outputs for EBV training, iBCE-EL outputs for EBV training, LBtope outputs for EBV training
2. Ground truth file needs to be in a separate folder and it's path entered manually to into line 10.
3. Enter path to where the final prediction** will be saved (line 76 in s_mv.py and line 71 in s_app.py)

1. All file paths to be entered manually into their respective variables
2. Weights to be set accordingly (line 32) (see weights below)
3. Enter path to where the final prediction** will be saved (line 90)

• EBV - Training [0.50, 0.51, 0.49, 0.54]

• EBV - Holdout [0.50, 0.51, 0.51, 0.54]

• HepC - Training [0.56, 0.56, 0.44, 0.71]

• HepC - Holdout [0.37, 0.61, 0.63, 0.73]

• Ovolvulus - Training [0.41, 0.55, 0.59, 0.55]

• Ovolvulus - Holdout [0.45, 0.54, 0.55, 0.55]

• Spyogenes - Training [0.60, 0.53, 0.41, 0.53]

• Spyogenes - Holdout [0.50, 0.50, 0.50, 0.80]

1. All file paths entered manually into their respective variables (Lines 9-25; Lines 66-78)
2. Comment/uncomment variables to accommodate your desired prediction (Lines 9-25; Lines 66-78)
3. Enter path to where the final prediction** will be saved (line 115)

** All final predictions are saved into a table containing the columns: [Info_protein_id], [Info_pos] necessary for the gather_results function in R