pab1 · Graph CNN

reference_pab1_gcn_plm_contiguousf3_20261105-040853

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	1.0539	0.8706	0.8765	0.7344	1489.0000	0.0139	4951.0000	0.1000	-0.1717	0.1945	0.7055	0.2600	-0.1638	0.3029	0.0000	0.7502	0.2700	-0.1574	0.5059	0.0200	0.7734	0.2850	-0.1460	0.5059	0.0467	0.8017
validation	0.9170	0.7577	0.6277	0.5084	931.0000	0.0404	1511.0000	0.2000	-0.2704	0.2331	0.3612	0.2200	-0.3375	0.8601	0.0400	0.3788	0.2300	-0.3475	0.8601	0.0200	0.4029	0.2000	-0.3642	0.8601	0.0444	0.3598
test	0.8171	0.8271	0.7070	0.6780	383.0000	-0.0396	1767.0000	0.1000	-0.3209	0.0138	0.4363	0.2400	-0.1896	0.4637	0.0400	0.5769	0.2300	-0.1733	0.4974	0.0900	0.6349	0.1562	-0.2219	0.4333	0.0000	0.5352

Loss per epoch

training validation

Config

name: reference_pab1_gcn_plm_contiguousf3
task: dms
data:
  dataset: pab1
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  graph_threshold: 7
  split_scheme: contiguous
  split_fold: 3
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