bgl3 · Graph CNN

reference_bgl3_gcn_mr5plm_modulof2_20261104-222701

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	0.6561	0.7756	0.7969	0.5884	143.0000	0.0573	62.0000	0.5000	-0.0727	0.3347	0.0000	0.6001	0.3600	-0.1118	0.3347	0.1400	0.7200	0.2900	-0.1069	1.6908	0.2100	0.7839	0.4444	-0.0640	0.3347	0.0741	0.6920
validation	0.7130	0.5405	0.5065	0.2890	44.0000	-0.1161	299.0000	0.1000	-0.8131	0.0490	0.0000	0.0690	0.1400	-0.6719	0.1561	0.1400	0.2233	0.1300	-0.6148	0.1847	0.2600	0.3067	0.1429	-0.6188	0.0490	0.0000	0.2038
test	0.9502	0.4780	0.5407	0.1912	23.0000	0.2723	1.0000	0.4000	-0.2878	0.3940	0.3000	0.5961	0.1400	-0.7297	0.3940	0.2000	0.3241	0.1000	-0.6083	0.3940	0.3000	0.4628	0.3750	-0.4004	0.3940	0.2500	0.5032

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_mr5plm_modulof2
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index,plm,plm_sum
  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