mseok / PIGNet2

MIT License
15 stars 8 forks source link

Output format #3

Closed sistar2020 closed 2 months ago

sistar2020 commented 3 months ago

Can you explain the columns in output files? For example, the content of examples/case1.txt is as below. Because this file does not have a header, it's hard to figure out what these columns represent.

$ cat examples/case1.txt
protein_ligand_single_conformation_0    0.000   -3.990  -2.074  -1.021  0.000   -0.894  0.000

My best guess is -3.990 is the total energy which corresponds to the sum of energy components (-2.074, -1.021, 0.000, -0.894 and 0.000). It will be appreciated if some comments are added on the output.

sangyeon-hits commented 3 months ago

Hello sistar2020,

Each of the numeric columns corresponds to:

  1. True label (which is just set to 0.000 in inference)
  2. Total predicted binding affinity (= sum of the right-hand values)
  3. van der Waals energy
  4. hydrogen bond energy
  5. metal-ligand coordination energy
  6. hydrophobic energy
  7. dummy variable (please ignore this column)