Fig. 4: Deep-learning models can represent RNA-structure-based observables.
From: Deep learning models for predicting RNA degradation via dual crowdsourcing
a, Representative structures from the best-predicted constructs from SHAPE modification (top row) and degradation at 10 mM Mg2+, pH 10, 1 day, 24 °C (Deg_Mg_pH10, bottom row). b, Nullrecurrent model predictions and experimental signal, averaged over secondary structure motifs. c, One failure mode for prediction came from constructs whose input secondary structure features were possibly incorrectly predicted.
