Prediction of RNA-RNA Interaction slides by Mathias M¨ ohl and Rolf Backofen ohl � M.M¨ c 1
RNA-RNA interaction (Waters and Storz, Cell 2009)� ohl � M.M¨ c 2
RNA-RNA interaction U U G U G C G C C C C U C G U U U U U C U G G G G C G U C G U U A C G U A C C C C C U C G U U U U A U A C G G A G C A C G G A G C A A U C A U U A G G G G A A G C A A U C A U G C A A A U U C A U U A C A U U A C C U A U U U U G G G G A A A A C C C C C C C C U U U U G G G G C C C C U U U U C C C C C C C C U U U U U U U U A A A A A A A A C C C C A A A A A A A A U A C G A U A C G A U A C G A U A C G A (Waters and Storz, Cell 2009)� ohl � M.M¨ c 3
How is a mRNA-target recognized? • idea 1: only hybridization energy counts U U G G C C C U C G U U A C G G A G C A A U C A U U A opt U G A C C U G C U C C U U A A C A A U A C G A • many approaches build on that: RNAhybrid, RNAplex, TargetRNA etc. • problem: structure U U A U G C A C U U A U G G G C A C C C U C G U U A U A U C G G A G C A A U C A U U A U versus G C G U G A C G U C C U G C U C C U U U A U A A A C C G C A G A C U A C U A C G A G C U G U C A A C U A C A C A • hence: additional information in accessibility of site ohl remark: accessibility = single-strandedness � M.M¨ c 4
Approaches to Target Detection approach 1 : maximize duplex energy (RNAhybrid, RNAplex, etc.) U U A U G C A C U U A U G G G C A C C C U C G U U A U A U C G G A G C A A U C A U U A U opt problem: G C G U G A C G U C C U G C U C C U U U A U A A C C G C A G A C U A C U A C G A G C U G U C A A C U A C A C A approach 2 : common structure by concatenation: (RNAcofold, PairFold) GGUGUUGGGAUUGUCAG • given: two sequences CUUACACAUCGGAGCAAUCAUUAGCUGUUCCUCAAUACGA • concatenate them GGUGUUGGGAUUGUCAG CUUACACAUCGGAGCAAUCAUUAGCUGUUCCUCAAUACGA • use Zuker’s algorithm G A C C A U U A U U U A G A U G U C C A A U A A U G A U C A G A C G A U problem: A U U C G C G C C G G A U G G U G U G G U C A G C A U A U U C A U A G ohl C U C A A � M.M¨ U U A G C A G C G U A G C A C U C A c A C A U C C U G U A 5
Approaches to Target Detection approach 3 : RNAup/IntaRNA: • determine probability for region i..j being unpaired • calculate ensemble energy from probability • hybridize unpaired region with second RNA Comparison: • approach 3: just one interaction possible • approach 2: more than one interaction possible, but only external interactions (no pseudoknots in concatenated structure) RNAup RNAcofold ohl � M.M¨ c 6
The Idea of IntaRNA IntaRNA = Int er a cting RNA similar to RNAup, but much faster (optimized for scanning genomes) mRNA i i’ k k’ ncRNA E hybrid ED mRNA ED ncRNA = + + E i , i ′ k , k ′ ↓ ↓ ↓ ohl as in RNAhybrid RNAplfold RNAplfold � M.M¨ c 7
Effificient Unpaired Probabilities (RNAplfold) Given RNA sequence S[1..n], compute probability Pr [ x .. yunpaired | S ] that positions x .. y of S are unpaired. Recall matrices of McCaskill: Q , Q b , Q m , Q m 1 , and introduce “outside” matrices ˆ Q , ˆ Q b Pr [ x .. y unpaired | S ] = Q u ( x , y ) / Q (1 , n ) Cases I x .. y external, O (1) II x .. y in hairpin, closed by ( i , j ), naive O ( n 2 ) III x .. y in internal loop, closed by ( i , j ), 5’ or 3’ of inner base pair (k,l), O (1) (internal loop size restricted) IV x .. y in multiloop, closed by ( i , j ), 5’, 3’, or between inner base pairs, naive O ( n 3 ) RNA Accessibility in cubic time Stephan H. Bernhart, Ulrike ohl � M.M¨ M¨ uckstein, Ivo L. Hofacker. AMB 2011 c 8
Two Parts of One Problem ? RNAup/IntaRNA RNAcofold - - - intern 1 2 extern Number of Interaction Sites Type of Interactions • however: there are more complex structures • double kissing hairpins • . . . ohl � M.M¨ c 9
Example • more than one internal interaction site • example: OxyS-fhlA interaction predicted complex [Alkan et al: JCB 2006] ohl � M.M¨ (Argamana and Altuviaa: JMB 2000) c 10
Generalized Problem approach 4 : predict joint mfe structure of two sequences • like Zuker on two sequences at the same time, including loops between the sequences • no pseudoknots, no crossing interaction • proven to be NP-complete • NP-completeness because of ZIG-ZAG structure • without ZIG-ZAGs polynomial algorithm ohl � M.M¨ c 11
Generalized Problem approach 4 : predict joint mfe structure of two sequences • like Zuker on two sequences at the same time, including loops between the sequences • no pseudoknots, no crossing interaction • proven to be NP-complete • NP-completeness because of ZIG-ZAG structure • without ZIG-ZAGs polynomial algorithm ohl � M.M¨ c 11
Efficient Base Pair Maximization without Zig-Zag Given sequences R and S , compute the maximal number of intramolecular and intermolecular base pairs Fragments/subproblems R [ i .. j ] , S [ k .. l ] Decomposition cases I unpaired base at either end of one sequence II closed structure at either end of one sequence III base pair enclosing either sequence IV interaction between left or right ends of sequences V decomposition at i ≤ i ′ ≤ j , k ≤ k ′ ≤ l into two subproblems R [ i .. i ′ ] , S [ k .. k ′ ] and R [ i ′ + 1 .. j ] , S [ k ′ + 1 .. l ] Complexity O ( n 6 ) time / O ( n 4 ) space. Why no zig-zags? ohl � M.M¨ c 12
RNA-RNA interaction: literature 1/2 approach 1: Hakim Tafer, Ivo L. Hofacker, RNAplex: a fast tool for RNA-RNA interaction search, Bioinformatics 2008 approach 2: Stephan H. Bernhart, Hakim Tafer, Ulrike M¨ uckstein, Christoph Flamm, Peter F. Stadler, Ivo L. Hofacker, Partition function and base pairing probabilities of RNA heterodimers, Algorithms Mol Biol 2006 ohl � M.M¨ c 13
RNA-RNA interaction: literature 2/2 approach 3: Ulrike M¨ uckstein, Hakim Tafer, Jorg Hackermuller, Stephan H. Bernhart, Peter F. Stadler, Ivo L. Hofacker, Thermodynamics of RNA-RNA binding, Bioinformatics 2006 Anke Busch, Andreas S. Richter, Rolf Backofen , IntaRNA: efficient prediction of bacterial sRNA targets incorporating target site accessibility and seed regions approach 4: Hamidreza Chitsaz, Raheleh Salari, S. Cenk Sahinalp, Rolf Backofen, A partition function algorithm for interacting nucleic acid strands, Bioinformatics 2009 Raheleh Salari, Mathias M¨ ohl, Sebastian Will, S. Cenk Sahinalp, Rolf Backofen, Time and space efficient RNA-RNA ohl interaction prediction via sparse folding, RECOMB 2010 � M.M¨ c 14
Recommend
More recommend
