id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-103853-ar09nzmw	Wayment-Steele, Hannah K.	RNA secondary structure packages ranked and improved by high-throughput experiments	2020-05-31		.txt	text/plain	4328	201	42	20 In this work, we evaluate the performance of commonly-used packages capable of making thermodynamic predictions in two tasks that have been crowdsourced on Eterna and are emerging as central to RNA characterization and design: 1) predicting chemical reactivity data through calculating probabilities that nucleotides are unpaired, and 2) predicting relative stabilities of multiple structural states that underlie the functions of riboswitch molecules, a task that involves predicting affinities of both small molecules and proteins of interest. We evaluated commonly used secondary structure modeling packages in their ability to make thermodynamic predictions on a compilation of large datasets of diverse synthetic molecules from Eterna, which we termed EternaBench ( Figure 1A ). We trained models with a variety of combinations of data types to explore interactions in multitask training (Table 1) and evaluated performance on held-out test sets for single-structure prediction accuracy, chemical mapping prediction accuracy, and riboswitch fold change prediction.	./cache/cord-103853-ar09nzmw.txt	./txt/cord-103853-ar09nzmw.txt