Skip to main content

Table 2 Training and validation datasets prepared from the loci in Table 1

From: A semi-supervised deep learning approach for predicting the functional effects of genomic non-coding variations

Cell lines

Training

Validation (pos/neg)

Labeled (pos/neg)

Unlabeled (pos/neg)

Lymphoblastoid (GM12878)

550 (275/275)

2606 (272/2,334)

272 (136/136)

Liver carcinoma (HepG2)

450 (225/225)

1297 (197/1,103)

208 (104/104)

Erythroleukemia (K562)

350 (175/175)

1210 (97/1,113)

136 (168/168)

  1. pos a positive locus affecting gene expression, neg a negative locus showing no effect