bgl3 · Graph CNN

reference_bgl3_gcn_mr5_modulof2_20261104-222113

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	0.5939	0.8085	0.8038	0.6274	143.0000	0.1123	107.0000	0.5000	0.0069	0.3347	0.0000	0.6376	0.3400	-0.1245	0.4008	0.1200	0.7126	0.3100	-0.1486	0.9971	0.2400	0.7576	0.3333	-0.0720	0.3347	0.0741	0.6877
validation	0.9650	0.2188	0.1977	0.0377	44.0000	-0.0178	322.0000	0.1000	-1.5074	0.2595	0.1000	-0.4483	0.0800	-0.9884	0.2595	0.1000	-0.0790	0.0700	-0.8720	0.3192	0.1700	0.0355	0.0000	-1.6200	-0.0191	0.0000	-0.5075
test	1.1547	0.2441	0.1897	0.0172	23.0000	0.0633	379.0000	0.0000	-1.8330	-0.2552	0.0000	-0.5308	0.0200	-1.5005	0.2838	0.0200	-0.3412	0.0200	-1.2506	0.2838	0.1100	-0.1354	0.0000	-1.7520	-0.2552	0.0000	-0.4618

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_mr5_modulof2
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index
  split_dir: null
  graph_threshold: 6
  split_scheme: modulo
  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