pab1 · Graph CNN

reference_pab1_gcn_contiguousf1_20261104-154441

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	2.0863	0.6617	0.6146	0.2756	2849.0000	0.0803	10159.0000	0.2000	-0.2493	0.4346	0.4471	0.2000	-0.2308	0.4346	0.0000	0.5058	0.2300	-0.2265	0.7565	0.0100	0.5346	0.2576	-0.1866	0.8601	0.0354	0.5849
validation	4.3330	-0.1565	-0.1454	-0.0202	475.0000	0.0843	3069.0000	0.0000	-5.4590	-3.1968	0.0000	-1.0911	0.0200	-4.6164	0.2486	0.0000	0.0000	-0.8634	0.0300	-4.2119	0.2486	0.0100	0.0000	-0.7404	0.0208	-4.5996	0.2486	0.0000	0.0000	-0.8556
test	3.6229	-0.1517	-0.1611	-0.0093	196.0000	0.0592	95.0000	0.0000	-2.4719	-0.0729	0.0000	-0.1691	0.0000	-2.7835	-0.0729	0.0000	0.0000	-0.3151	0.0200	-3.0178	0.6164	0.0200	0.0000	-0.4294	0.0000	-2.5602	-0.0729	0.0000	0.0000	-0.2123

Loss per epoch

training validation

Config

name: reference_pab1_gcn_contiguousf1
task: dms
data:
  dataset: pab1
  encoding: one_hot,aa_index
  split_dir: null
  graph_threshold: 7
  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