RNAMoIP

Parameters

Sequence:

A valid RNA sequence as a case-unsensitive IUPAC-like format. This will be the base of the prediction.

Structure (Optional):

A valid RNA structure that will be take as constraints for the base pairings probabilities prediction. Pairing levels must be correctly identified, with the following syntax (show ordered by the pairing level): .()[]{}<>

Alpha:

Sets how the RNAMoIP optimize its objectives

Max level:

The maximum depth level of base pairs crossing allowed.

Max Solution Count:

The maximum number of best solutions return by the IP solver.

Information on the alpha parameter

RNAMoIP exposed an alpha parameter that have some agency on what the objective function should maximise. Lets take a prediction for PDB 5BTP-B as an example. The parameters will look like the following:

For a prediction with alpha=0, RNAMoIP will only look to maximise the base pairing probabilities inserted. This means that motif insertion will be ignored. For a prediction with alpha=1, RNAMoIP will only look to maximise the length of the motifs inserted. This means that base pair will only be inserted to enable some motifs. Note that our motif also defines possible interactions inside them, meaning the prediction is actually meaningful. For anything in between, RNAMoIP will consider both base pairing probabilities and motifs. In our research, we found that an alpha of 0.1 yield the best result, where the base pairs inserted will guide where the motifs should also be located. Here are the results summarized:

Alpha	Predicted Secondary Structure	VARNA Representation
Real Structure	........((((....[[[[[...((((......)))).....))))............(((..]]]]]..))).
Alpha = 0	[[(((((([[[[))))))(((...((((......)))).....]]]].(((.]]....[[[[...))))))]]]]
Alpha = 0.1	.....((((.((....[[[[....((((......)))).....)).))))........(((....]]]]...)))
Alpha = 1	.....((..........((......(((......)))......))...))........(((...........)))