bgl3 · Graph CNN

reference_bgl3_gcn_plm_contiguousf4_20261105-043256

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	1.8359	0.2727	0.2328	0.0676	5893.0000	-0.0101	10364.0000	0.5000	0.5039	1.9459	0.2483	0.5000	0.2552	2.2610	0.0000	0.1925	0.6100	0.4082	2.4113	0.0100	0.2571	0.5987	0.4007	2.4113	0.0197	0.2630
validation	1.8043	0.3223	0.3007	0.0996	1218.0000	-0.0483	635.0000	0.5000	0.1296	1.3581	0.2302	0.3800	-0.1490	1.9459	0.0000	0.1776	0.3300	-0.2152	2.5819	0.0400	0.1674	0.3953	-0.1407	1.9459	0.0000	0.1777
test	1.7766	0.2636	0.2006	-0.0003	1997.0000	-0.1333	50.0000	0.3000	-0.6100	1.1624	-0.1420	0.4200	-0.1076	3.1160	0.0200	0.0316	0.4300	-0.0757	3.1160	0.0100	0.0470	0.3830	-0.2243	1.6946	0.0000	-0.0136

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_plm_contiguousf4
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  graph_threshold: 6
  split_scheme: contiguous
  split_fold: 4
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