Computational Molecular Microbiology (MBIO 4700) Lecture Notes PDF

Computational Molecular Microbiology (MBIO 4700) ABDULLAH ZUBAER UNIVERSITY OF MANITOBA Protein structure: physical methods X-ray crystallography NMR spectroscopy Cryo-electronmicroscopy Protein structure: modeling methods ▪ Homology modeling ▪ Fold recognition ▪ Ab initio Protein Folding...

Computational Molecular Microbiology (MBIO 4700) ABDULLAH ZUBAER UNIVERSITY OF MANITOBA Protein structure: physical methods X-ray crystallography NMR spectroscopy Cryo-electronmicroscopy Protein structure: modeling methods ▪ Homology modeling ▪ Fold recognition ▪ Ab initio Protein Folding Pevzner 2015 Protein structure: modeling methods ▪ Homology modeling: Swiss-Model, Modeller, Phyre2 ▪ Fold recognition: I-TASSER ▪ Ab initio : Rosetta https://zhanglab.ccmb.med.umich.edu/I-TASSER/ I-TASSER Iterative Threading ASSEmbly Refinement https://zhanglab.ccmb.med.umich.edu/I-TASSER/download/ Online server: https://zhanglab.ccmb.med.umich.edu/I-TASSER/ J Yang, R Yan, A Roy, D Xu, J Poisson, Y Zhang. The I-TASSER Suite: Protein structure and function prediction. Nature Methods, 2015, 12: 7-8. I-TASSER pipeline Sorting Similar structures “blastp” C N Ab initio protein modeling Robetta protein folding https://robetta.bakerlab.org/ Rossetta protein folding platform Can do de novo prediction OR Ab initio modelling NEED to set up an Account: Register Example: Strategy Assumption: protein folding not random -many conformations; need to consider: ◦ local secondary structures, ionic interactions, hydrophobic interactions, etc. (even chaperons) -Rosetta “ tries to mimic” steps involved in protein synthesis (ribosome) as the nascent peptides emerges and amino acids can interact within one another; fragments are selected from known structures; -short (3 or 9-mers) peptides are folded and compared to known folds (3 or 9mer libraries) and those folds are clustered according to energy values (energy based clustering) -analysis generates a large library of “short peptide folds” Overview: Known structures (possible conformations) Unknown (query) Generate fragment library (all possible fold for a 9 or 3 mer) Assemble (predict) structures (many combinations – need to sample and evaluate and resample etc.) http://robetta.bakerlab.org/ Based on known folds for short (3, 9) aa peptides Assembly of “short” peptides in various combinations to achieve optimal (low) energy values Many possible fold – screening Based on energy values and possible constrains (if provided) Reality: will get several “models” Models? May need other Tools to infer best Model (see notes on DALI etc.) From 2009. Protein Structure to Function with Bioinformatics; Li et al. Ab Initio Protein Structure Prediction (Chapter 1). Overview of generating folds: “Folding funnel” Strategy Time consuming (can take several months etc.) How good are the predictions: ◦ Based on energy values (should be in the range from RSE -100 to -500, Rosetta Energy units (RSE) ◦ Find similar structures (use search tools that use MC) ◦ Does it look like a “reasonable protein” – size/ shape etc. – do you have any ideas about plausible function? Does the “fold” make sense ? ◦ Fold recognition servers: DALI or PDBefold MORE updates (FYI): (PNAS 2020; trRosetta) Levinthal’s Paradox Melkikh AV, Meijer DKF. 2018. On a generalized Levinthal's paradox: The role of long- and short range interactions in complex bio-molecular reactions, including protein and DNA folding. Prog Biophys Mol Biol. 132:57-79. doi: 10.1016/j.pbiomolbio.2017.09.018. Unfolded Folding intermediates Most stable fold (biologically relevant fold) Chaperons – help in avoiding “folding traps” ! (hard to “model” the actions of chaperons) Is your protein an IDP? Intrinsic disordered proteins (IDPs) Or Intrinsic disordered regions (within proteins) (IDRs) http://bioinf.cs.ucl.ac.uk/intro duction/ http://bioinf.cs.ucl.ac.uk/psipred/ http://bioinf.cs.ucl.ac.uk/web_servers/ Protein Disorder (predictions) Dali server (helsinki.fi) 3D coordinates for your query And find Similar “shaped” Proteins Liisa Holm; Laura M. Laakso (2016) Dali server update. Nucleic acids research 44 (W1), W351-W355. PDBe < Fold < EMBL-EBI Comparing 3D structures! Other tools: SUPER Super · bio.tools Checks for “motifs”/folds that might be found in unrelated proteins. DeepView (ExPASy) Comparing structures Swiss PDB Viewer - Home (vital-it.ch) Classic IDPs Linker regions and signaling proteins (switches – requiring conformational changes); proteins that need to wrap around other proteins (dehydrants) or biomolecules; proteins that need to be dynamic in interacting with other proteins (nuclear pore complex proteins*) etc. General features of IDRs (IDPs): ◦ Rich in Prolines (helix breakers) ◦ Rich in glycines (short side chains – so flexible) ◦ Few hydrophobic amino acids ◦ Highly charged and repetitive *Hough LE, Dutta K, Sparks S, Temel DB, Kamal A, Tetenbaum-Novatt J, Rout MP, Cowburn D. 2015. Elife. 4. pii: e10027. doi: 10.7554/eLife.10027. Notes on Protein Disorder (IDR and IDP) KEYPOINT: Both structured and disordered regions are required for protein function. Conformational changes can relate to function and regulation. Babu et al. 2012. Science. 337: 1460 – 1461. Strategies for IDPs Get some biophysical data (via NMR etc.), allows for some parameter (restraints) estimations Rosetta – modelling If possible, design experiments to evaluate protein function Pevzner 2015