Reliable Prediction of Viral RNA Structures Winterseminar 2016, Bled Roman Ochsenreiter TBI Wien Bled, 15.2.2016 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 1 / 20
Viruses Figure: Crystal Structure and Schematic of Rhinovirus A 1 Genome size (nt): Human: 3.2 10 9 Rhinovirus: 7 10 3 1 Zhao et al. 1996 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 2 / 20
Viruses and RNA Structure Viruses have little space (small genomes) for information storage Specific Functions are often provided by RNA Structures Figure: Pestivirus IRES 1, Rhinovirus CRE-Element 1 Rfam IDs: RF00209, RF00220 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 3 / 20
Our Arsenal Tools for prediction of RNA Secondary Structure: Structure Prediction Tools RNAfold Prediction of MFE structures/structure Ensembles RNAalifold RNA folding with evolutionary constraints RNAplfold Local structure Prediction AliDot Heuristic - predicts plausible Structures from alignments RNAz Uses machine learning for conserved structure detection How do they perform on Viruses? Can we use our tools to find annotated structures? What is the optimal approach? Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 4 / 20
HIV1 Structure - The Gold Standard Figure: HIV1 genome, folded with SHAPE-Data 1 1 Watts et al. 2009 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 5 / 20
Local vs. Global Folding How far reaching are base-pairings in HIV1? Basepair Spans in HIV1 Counts 0 200 400 600 800 1000 1200 0 2000 4000 6000 8000 Basepair Span Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 6 / 20
Local vs. Global Folding How many HIV1-MFE basepairs do we miss by folding locally? Missed MFE basepairs in local Structures % 20 25 30 35 40 200 400 600 700 RNAplfold Window Size Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 7 / 20
Global vs. Local Folding RNAfold vs RNAplfold - Shared Basepairs RNAfold - with SHAPE constraints 1241 1225 1360 1105 Winsize 150 1124 1078 Winsize 300 2183 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 8 / 20
Local vs. Global Folding Missed MFE basepairs > Window size % 30 40 50 60 70 200 400 600 700 RNAplfold Window Size Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 9 / 20
Annotated viral RNA Structures Name Hits Length RRE 200 336 HIV GSL3 200 83 HIV POL-1 SL 200 112 HIV FE 186 51 HIV-1 SL4 175 19 HIV-1 DIS 148 39 mir-tar 139 60 HIV-1 SL3 129 22 HIV-1 SD 121 18 HIV PBS 90 96 astro FSE 3 51 mir-224 1 96 Table: Rfam - Annotated RNA structures in 200 Reference HIV1 genomes Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 10 / 20
Reproduction of Rfam-annotated Structures Element 1: HIV-1 DIS (Retroviral Psi packaging element) Figure: HIV Retroviral Psi packaging Element Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 11 / 20
Reproduction of Rfam-annotated Structures Reference (Rfam)...((((.(((((((...)))))))...))))... RNAfold (global) ))..)))(((((((((.(((((...((...) RNAfold (local)..((((.(((((((.(...).)))))))...))))... Table: HIV1 Psi Packaging Element, Alignment Size: 200, MPI = 86.91 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 12 / 20
Reproduction of Rfam-annotated Structures Reference (Rfam)...((((.(((((((...)))))))...))))... RNAfold (global) ))..)))(((((((((.(((((...((...) RNAfold (local)..((((.(((((((.(...).)))))))...))))... Alifold (local, clustal)..((((.(((((.((...)))))))...))))... Alifold (local, mlocarna)...(((((((...)))))))... Alifold (global) )).((((.(((((((...)))))))...))))... Table: HIV1 Psi Packaging Element, Alignment Size: 200, MPI = 86.91 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 13 / 20
Reproduction of Rfam-annotated Structures Reference (Rfam)...((((.(((((((...)))))))...))))... RNAfold (global) ))..)))(((((((((.(((((...((...) RNAfold (local)..((((.(((((((.(...).)))))))...))))... Alifold (local, clustal)..((((.(((((.((...)))))))...))))... Alifold (local, mlocarna)...(((((((...)))))))... Alifold (global) )).((((.(((((((...)))))))...))))... mlocarna..((((.(((((((...(...).)))))))...))).).. AliDot..((((...(((...)))...)))).. RNAz..((((.(((((.((...)))))))...))))... Table: HIV1 Psi Packaging Element, Alignment Size: 200, MPI = 86.91 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 14 / 20
Reproduction of Rfam-annotated Structures Element 2: HIV POL-1 SL (HIV pol-1 Stem Loop) Figure: HIV pol-1 stem loop Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 15 / 20
Reproducibility of HIV POL-1 SL Reference (Rfam)...(((((((((((((((((((...(((((((((...)))))))))...((((((...)))))).)))))))))))))))))))... RNAfold (global) )))(((((...((((.((((...(((((((((.((...)).)))))))))...)))).))))...((((((((((.(((((((...((((...))) RNAfold (local)...((((((((.((((((((((...(((((((((.((...)).)))))))))...((((((...)))))).)))))))))).)))))))).. Table: HIV1 Polymerase 1 stem loop, Alignment Size: 200, MPI=87.99 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 16 / 20
Reproducibility of HIV POL-1 SL Reference (Rfam)...(((((((((((((((((((...(((((((((...)))))))))...((((((...)))))).)))))))))))))))))))... RNAfold (global) )))(((((...((((.((((...(((((((((.((...)).)))))))))...)))).))))...((((((((((.(((((((...((((...))) RNAfold (local)...((((((((.((((((((((...(((((((((.((...)).)))))))))...((((((...)))))).)))))))))).)))))))).. Alifold (local, clustal)...((((...))))... Alifold (local, mlocarna)..((((((((.((((((((((...(((((((...((...))...)))))))...((...))...)))))))))).))))))))... Alifold (global)..((((((((.((((((((((...(((((((..(((...))).)))))))...(((.((...)).))).)))))))))).))))))))... Table: HIV1 Polymerase 1 stem loop, Alignment Size: 200, MPI=87.99 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 17 / 20
Reproducibility of HIV POL-1 SL Reference (Rfam)...(((((((((((((((((((...(((((((((...)))))))))...((((((...)))))).)))))))))))))))))))... RNAfold (global) )))(((((...((((.((((...(((((((((.((...)).)))))))))...)))).))))...((((((((((.(((((((...((((...))) RNAfold (local)...((((((((.((((((((((...(((((((((.((...)).)))))))))...((((((...)))))).)))))))))).)))))))).. Alifold (local, clustal)...((((...))))... Alifold (local, mlocarna)..((((((((.((((((((((...(((((((...((...))...)))))))...((...))...)))))))))).))))))))... Alifold (global)..((((((((.((((((((((...(((((((..(((...))).)))))))...(((.((...)).))).)))))))))).))))))))... mlocarna..((((((((.((((((((((...(((((((...((...))...)))))))...((...))...)))))))))).))))))))... AliDot... RNAz...((((...))))... Table: HIV1 Polymerase 1 stem loop, Alignment Size: 200, MPI=87.99 Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 18 / 20
Summary Even though long range interactions are found, the HIV1 genome is dominated by short-range base pairings Unconstrained, global Folding fails to reproduce known structures Sequence - based alignments can bias consensus-structure based folding of viral structures Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 19 / 20
Acknowledgements Ivo Hofacker Christoph Flamm TBI collegues..and you! Roman Ochsenreiter TBI Wien Reliable Prediction of Viral RNA Structures Bled, 15.2.2016 20 / 20