bgl3 · Graph CNN

reference_bgl3_gcn_mr5plm_contiguousf2_20261104-222732

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	1.1637	0.5944	0.6238	0.3417	152.0000	-0.0484	957.0000	0.1000	-0.9514	0.1164	0.0000	0.2825	0.0600	-0.5926	0.1164	0.0000	0.5137	0.1300	-0.6399	0.2019	0.0700	0.5191	0.0625	-0.7514	0.1164	0.0000	0.4131
validation	1.3683	0.5218	0.5042	0.2428	38.0000	-0.0084	251.0000	0.3000	-1.5690	0.3961	0.1000	-0.0215	0.1200	-1.4312	0.7630	0.1200	0.0532	0.0800	-1.3528	0.7630	0.1300	0.1080	0.2727	-1.6053	0.3961	0.0909	-0.0388
test	1.9258	0.3631	0.3163	-0.0103	26.0000	-0.1330	672.0000	0.0000	-1.7891	-0.3805	0.0000	-0.0175	0.0000	-1.9637	-0.1352	0.0000	-0.1161	0.0000	-1.7644	-0.1352	0.0400	-0.0068	0.0000	-1.6891	-0.2397	0.0000	0.0311

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_mr5plm_contiguousf2
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  graph_threshold: 6
  split_scheme: contiguous
  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