pab1 · Graph CNN

reference_pab1_gcn_plm_contiguousf1_20261104-181959

Metrics

Set	mse	pearsonr	spearmanr	r2	above_wt_n	above_wt_spearmanr	budget_to_max	hit_rate_at_10	mean_at_10	max_at_10	captured_at_10	hit_rate_at_50	mean_at_50	max_at_50	precision_at_50	captured_at_50	hit_rate_at_100	mean_at_100	max_at_100	precision_at_100	captured_at_100	hit_rate_at_1pct	mean_at_1pct	max_at_1pct	precision_at_1pct	captured_at_1pct
train	0.5535	0.9098	0.8567	0.8078	2849.0000	0.0352	293.0000	0.4000	-0.0535	0.2612	0.5221	0.5400	0.0303	0.5294	0.0000	0.6171	0.5400	0.0255	0.6318	0.0000	0.6478	0.5152	0.0187	1.1615	0.0505	0.6825
validation	1.7030	0.7802	0.7933	0.5990	475.0000	-0.0532	1251.0000	0.3000	-0.1320	0.0738	0.7283	0.2200	-0.1126	0.5226	0.0600	0.7898	0.2300	-0.1335	0.5226	0.0400	0.8101	0.2083	-0.1203	0.5226	0.0625	0.7855
test	1.8129	0.7334	0.7076	0.4949	196.0000	0.0632	1557.0000	0.1000	-0.3171	0.0916	0.6785	0.1600	-0.4177	0.3256	0.0400	0.6902	0.2200	-0.4055	0.3256	0.0900	0.7201	0.0833	-0.3964	0.2232	0.0000	0.6740

Loss per epoch

training validation

Config

name: reference_pab1_gcn_plm_contiguousf1
task: dms
data:
  dataset: pab1
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  graph_threshold: 7
  split_scheme: contiguous
  split_fold: 1
model:
  name: gcn
  filters:
  - 32
  hidden:
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 128
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7