Table 1.

Comparison of commonly used methods for lncRNA structural studies.

Type of Experiment	Limitations		Methods	Structure	lncRNA	References
In vitro	Controlled conditions, lncRNA structure more stable than in in vivo tests Requires sequencing		Enzymatic probing
		Only in vitro Fragmentation of RNA by nucleases makes difficult the identification of secondary structures at single-nucleotide level Might require additional probing with nuclease or chemical	PARS/nuclease S1 and RNaseV1	double- and single- stranded regions of RNA	Rox1 Rox2	[73]
			Chemical probing
		Does not provide information concerning interactions between base-pairing (at close- or long-range) Labour-intensive	SHAPE-seq	2′-OH acylation	Braveheart RepA Rox1 Rox2 SRA HOTAIR COOLAIR MALAT1 NEAT1	[66,67,68,69,71,73,74,75]
		DMS- nucleotide bias (only able to react with purines) DMS is corrosive and toxic, In some cases, using DMS is not sufficient to capture all single-stranded regions Labour-intensive	DMS-seq (DMS)	unpaired adenine and cytosine residues	BraveheartRepA SRA HOTAIR MALAT1	[66,67,69,74,75]
In vivo	Requires sequencing Difficult task due to lncRNAs’ size and low abundance in cells lncRNAs are expressed in alternative isoforms and bound by a variety of RNA binding proteins in vivo		Chemical Probing:
			SHAPE-MaP (1M7,1M6,NMI1)	2′-OH acylation	Xist	[55]
In silico	lncRNA is more structured than in in vivo tests Only for prediction of secondary structure Has to be complemented by in vivo/in vitro tests		CROSS (Computational Recognition of Secondary Structure)	RepA, D2 domain	Xist HOTAIR	[76]
Biophysical	Difficult for application of higher numbers of transcripts Difficult for long RNA strands Requires large quantities of robust, homogenous sample Crystallography is labour-intensive		X-ray	A-rich 3′-UTR	MALAT1	[77]
			NMR spectroscopy	AUCG tetraloop	Xist	[56]