Skip to main content

Table 1 SsNA secondary structure prediction tools included in the present study

From: Comparative Study of Single-stranded Oligonucleotides Secondary Structure Prediction Tools

Method

Year

Prediction approach

ssDNA parameter

Pseudoknots

Speed

Mfold

2003

MFE-based

Yes

No

\(O(n^3)\)

RNAfold

2008

MFE-based

Yes

No

\(O(n^3)\)

CentroidFold

2009

MFE-based or knowledge-based

with \(\gamma\)-centroid estimator

No

No

\(O(n^3)\)

LinearFold

2019

MFE-based or knowledge-based

with runtime linearization and heuristic beam search

No

No

O(n)

CONTRAfold

2006

Machine learning

No

No

\(O(n^3)\)

MC-fold

2008

Machine learning

No

Yes

\(O(n^{{}^{15}/{}_{2}})\)

MXfold2

2021

Deep learning & thermodynamic

No

No

0.31 s (GPU)a

UFold

2022

Deep learning

No

Yes

0.16 s (GPU)a

SPOT-RNA

2019

Deep learning

No

Yes

77.80 s (GPU)a

  1. aAverage run-time measured for sequences of 1000–1500 nucleotides