bgl3 · Graph CNN

reference_bgl3_gcn_mr5_randomf2_20261104-222012

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	precision_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	0.5030	0.8571	0.8554	0.7298	124.0000	0.1800	2.0000	0.4000	0.0617	0.9971	0.2000	0.7392	0.3200	-0.0476	0.9971	0.1800	0.8043	0.3100	-0.0564	0.9971	0.2800	0.8557	0.3824	0.0064	0.9971	0.2059	0.7993
validation	1.0158	0.6800	0.6968	0.4593	40.0000	-0.2316	156.0000	0.0000	-0.6106	-0.0741	0.0000	0.5297	0.1400	-0.5557	0.0925	0.1600	0.6275	0.1200	-0.4643	0.2595	0.2400	0.7185	0.0000	-0.5822	-0.0741	0.0000	0.5468
test	0.9574	0.7043	0.7183	0.4934	53.0000	-0.2098	14.0000	0.5000	-0.4038	0.5199	0.1000	0.5736	0.2600	-0.4065	1.6908	0.2400	0.6815	0.1900	-0.4603	1.6908	0.3000	0.6967	0.4545	-0.4770	0.5199	0.0909	0.5468

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_mr5_randomf2
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index
  split_dir: null
  graph_threshold: 6
  split_scheme: random
  split_fold: 2
  min_reads: 5
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