Extended Data Fig. 5: Benchmarks for training, validation, and testing HORNET.
From: Determining structures of RNA conformers using AFM and deep neural networks
Percentages of the total simulated data used for training (pseudo-structure database: 80% BM0, 5% BM1, 5% BM2), training-validation (20% BM0, 1% BM1, 1% BM2), validation (5% BM5), blind testing (100% BM3 and BM4) and testing (all remaining data).
