Informatica Biomedica

lezione18

Alberto Paoluzzi Mauro Ceccantiwww.dia.uniroma3.it/ paoluzzi/web/did/biomed/

Informatica e Automazione, "Roma Tre" — Medicina Clinica, "La Sapienza"

May 17, 2010

Informatica Biomedica: Lezione 18

Molecular Visualization

α-shapes

Fonte essenziale: Chapter 9 of book: J. Gu and P.E. Bourne,Structural Bioinformatics, Wiley (2009)

Sir John Kendrew with the model of insulin.

The model ofinsulin was one ofthe first proteinstructures to bedetermined byX-raycrystallography

Prediction of threedimensional structure

The geneticists now believe that the hereditary materialdetermines only the amino acid sequence of a protein, notits 3D structure. That is to say, the polypeptide chain,once synthesized, should be capable of folding itself upwithout being provided with additional information. If thepostulate is true it follows that one should be able topredict the threedimensional structure of a protein from aknowledge of its amino acid sequence alone.Indeed, in the very long run, it should only be necessaryto determine the amino acid sequence of a protein, andits 3D structure could then be predicted; in my view thisday will not come soon, [] and it will also be possible todiscuss the structures of many important proteins whichcannot be crystallized and therefore lie outside thecrystallographer’s purview.

Myoglobin and the structure of proteins. (1962) Nobel Lecture– John C. Kendrew

A typical fragment of electron density and a section of

atomic model

the structure ofthe CuA domainfrom cytochromeBC3 (PDB IDcode 2CUA),displayed usingXFit fromXtalView package.Bonds are coloredaccording to theatoms that theyjoin. Putativehydrogen bondsare drawn asdashed white lines.

Visualization of surface (globular protein 8GCH)

PMV fromMGLTools is apowerful molecularviewer that has anumber ofcustomizablefeatures andcomes with manypluggablecommands rangingfrom displayingmolecular surfacesto advancedvolume rendering.

A region of human rhinovirus 1A (HRV-1A)

� including a bounddrug molecule(PDBid 2HWD).

� The virus proteinsare shown as asimple backbonetrace, with the drugrepresented asspace-fillingspheres.

Human rhinovirus 1a coat protein Human rhinovirus 1a coat protein

CATH classificationDomain

2hwd100

ClassMainly Beta

ArchitectureSandwich

TopologyJelly Rolls

Secondary structure of human thioredoxin (1ERT)

� The protein β-strands arerepresented by arrowsfrom the N- to theC-terminus, and α-helicesas spiral ribbons.

� Regions without definedsecondary structure areshown as a smooth tube.

� The four β-strands form aβ-sheet at the center ofthe structure, easily visiblein this kind ofrepresentation.

� Image generated byMolScript and render.

A molecular surface drawn as a mesh.

� A molecular surfacedrawn as a mesh,

� overlaid on asecondary structurerepresentation ofthe toxin LQ2 fromLeiurusQuinquestriatus(PDBid 1LIR).

� image was preparedwithin PyMol

Structure of c-AMP-Dependent protein kinase (2CPK)

� displayed usingWebMol andMolSurfer (Java)

� some fairlyadvanced features:

1. interactiveRamachandranplot

2. link of a 2Dprojection of amacromolecularinterface to a3D view of thestructures.

MICE scene interchange format

The MICE project is addressing these issues through developmentof a custom XML-based language.The Molecular Scene Description Language (MSDL) will provide aflexible system for describing the appearance of a molecularstructure, in a platform- and application-independent fashion.Need of tools for analyzing, understanding, and exchanging

data

The volume of data produced by genome projects, X-raycristallography, NMR spectroscopy, and electron andconfocal microscopy presents the bioinformaticscommunity with new challenges for analyzing,understanding, and exchanging this data

Molecules to Maps: tools for visualization and interaction insupport of computational biology

– Eileen T. Kraemer and Thomas E. Ferrin

Bioinformatics review, 14-9, 19998

Frequently, data in scientific computing is in its abstract form afinite point set in space, and it is sometimes useful or required tocompute what one might call the “shape” of the set.

The article

Edelsbrunner, H. and Mücke, E. P. 1994. Three-dimensional alphashapes. ACM Trans. Graph. 13, 1 (Jan. 1994), 43-72. DOI=http://doi.acm.org/10.1145/174462.156635

introduces the formal notion of the family of α-shapes of a finitepoint set in R3.

Each shape is a well-defined polytope, derived from the Delaunaytriangulation of the point set, with a parameter α ∈ R controllingthe desired level of detail.

An algorithm is presented that constructs the entire family ofshapes for a given set of size n in time O( n2),worstcase.

aaaaaaa

aaaaaaa aaaaaaa

aaaaaaa aaaaaaa