bgl3 · Graph CNN

reference_bgl3_gcn_contiguousf1_20261104-154024

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	1.7888	0.4656	0.4642	0.0856	5657.0000	0.0782	5036.0000	0.5000	0.4152	2.3637	0.2198	0.4800	0.0540	2.3637	0.0000	0.1317	0.5400	0.1936	2.5419	0.0200	0.1886	0.5533	0.2010	2.5419	0.0267	0.1985
validation	2.1369	-0.0562	-0.0911	-0.0009	1465.0000	-0.0232	2881.0000	0.3000	-0.1612	1.3581	0.0000	0.1027	0.4600	0.0410	2.4113	0.0200	0.0000	0.1870	0.3700	-0.5112	2.4113	0.0200	0.0000	-0.0063	0.4884	0.1839	2.4113	0.0233	0.0000	0.2335
test	1.6897	-0.0871	-0.0962	-0.0200	1203.0000	-0.0299	2631.0000	0.0000	-2.5489	-0.5615	0.0000	-0.8091	0.1000	-1.8849	1.3581	0.0000	0.0000	-0.6207	0.1300	-1.6617	1.3581	0.0000	0.0000	-0.5714	0.0882	-1.8335	1.2150	0.0000	0.0000	-0.5822

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_contiguousf1
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index
  split_dir: null
  graph_threshold: 6
  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