1	Methods for Receptors with Known Sequence
2
3	Genomic Information three most comprehensive data banks --- GenBank www.ncbi.nih.gov/Genbank/ --- EMBL-EBI Nucleotide Sequence Database www.ebi.ac.uk/embl/ --- DNA Data Bank of Japan www.ddbj.nig.ac.jp daily exchange of data tools for processing the information
4	DNA Sequence Types in Databanks mRNA - may contain introns (non-coding sequences) and may not cover complete coding region cDNA - derived from mRNA by reverse transcription genomic DNA - from genome sequencing project; may contain introns, repeat regions, and other features; usually complete GSS (Genome Survey Sequence) - single-pass sequence with many errors EST (Expressed Sequence Tags) - short cDNA sequences prepared from mRNA from cell under particular conditions (e.g. disease); will not cover complete coding region.
5	Target Sequences for Drug Design proteome mapping in disease --- the sequences that are over-expressed can be targeted protein-protein interactions – micro-arrays --- interactions that are occurring in disease can be prevented by compounds that bind to one of the proteins in the binding area
6	From Genome to Proteins the aim for drug design: DNAàRNAàproteinà3D-structureàbinding sequences of nucleotides (A, C, G, T) available, the aim is to translate them to amino acid sequences coding regions for proteins (genes/cistrons) – no straightforward to find, annotation (sequence features) important amino acids coded by codons (triplets) but the sequence also contains start and stop codons…
7	Translation Table(s)
8	Translation DNA contains two complementary antiparallel strands 5’-AGCAGTCGATGCCGAATTCC-3’ 3’-TCGTCGACTACGGCTTAAGG-5’ each strand can be read in two directions for each direction, there are three possible ways – six possible reading frames those that have stop codons early in the sequence can be discarded
9	Protein Sequence Databases I
10	Protein Sequence Databases II
11	Protein Sequence Databases III
12	Post-translational Modifications
13	Sequence Alignment I structure is more conserved than sequence --- similarity in sequence implies similarity in structure --- proteins with different sequences can have similar structures similarity in sequence also implies similarity in function comparison of the same protein in different organisms --- evolutionary relationships (phylogenetic analysis) when doing alignments, one can look at --- identical amino acids --- amino acids with similar properties
14	Sequence Alignment II sequence identity --- the same amino acids at equivalent positions sequence similarity --- amino acids with similar properties can be interchanged sequence homology --- homology and similarity are often used as synonyms --- in phylogenetic analysis, two sequences are homologous if they share a common ancestor
15
16	Sequence Alignment: Procedure the compared sequences are --- moved along each other --- the gaps are introduced if needed, penalties - for opening a gap - for extending a gap (smaller) similarity matrix --- rows and columns are individual positions --- elements contain scores for comparison - highest scores for identity - lower scores for similarity - negative scores for dissimilar substitutions the best alignment has the highest score
17	Sequence Alignment: Dotplots the plot of the relation between two sequences --- the dots mark positions of identical AA --- can be optimized by ‘sliding window’
18
19	Sequence Alignment: Two Sequences lalign --- www.ch.embnet.org/software/LALIGN_form.html SIM --- ca.expasy.org/tools/sim.html align --- www.ebi.ac.uk/emboss/align/ the best way to align two sequences is to align the whole family --- varying AAs less important --- conserved AAs functionally important --- gaps frequently in loops
20	Sequence Alignment: Databases Basic Local Alignment Search Tool (BLAST) --- www.ncbi.nlm.nih.gov/BLAST/ --- uses local alignments --- able to detect relationships among sequences which share only isolated regions of similarity --- various enhancements and variants available FASTA --- www.ebi.ac.uk/fasta33/ ClustalW --- www.ch.embnet.org/software/ClustalW.html T-COFFEE (very good, comparatively slow) www.ch.embnet.org/software/TCoffee.html
21	Sequence Alignment: Quality % identity Expect (E-) value --- takes into account the size of the database --- describes the number of hits one can "expect" to see just by chance when searching a database of a particular size E-value – rule of thumb --- E < 0.1 good, random alignment --- E ~ 1 means 50% chance alignment --- E > 10 no relationship
22	Motifs specific consensus patterns discovered in multiple alignments discovered through profiles, Hidden Markov Models, position specific score matrices indicative of protein function --- function can be identified even if overall similarity is low organized in motif databases many motif databases can be searched simultaneously by InterPro --- www.ebi.ac.uk/interpro/scan.html
23	Alignment Plus Motifs Position-specific Iterated BLAST (PSI-BLAST) --- www.ncbi.nlm.nih.gov/BLAST/ initial alignments generate a profile --- frequencies of AAs in individual positions motifs can be incorporated easily the procedure is iterated until no statistically different sequences are found in the database very good alignment tool
24
25	From Sequence to Structure
26	What Is a Template?
27	Comparative Modeling identify and select related structures (templates) align target sequence with templates --- for drug design, similarity >75% --- gaps should be between secondary structure elements build a structural model for the target using information about template structures --- in principle, copy template structure to target --- gaps – loop search algorithms for structures from PDB-based libraries or optimization evaluate model (repeat if not satisfied) --- procheck (bonds, dihedrals, noncovalent int’s…) --- molecular modeling (side chain optimization, MD)
28	Comparative Modeling: Software Modeller (Andrej Sali) – unix software MODBASE (http://guitar.rockefeller.edu/modbase) is a relational database of annotated comparative protein structure models for all available protein sequences matched to at least one known protein structure SWISS-MODEL – an easy to use web-server --- www.expasy.ch/swissmod/SWISS-MODEL.html
29	Comparative Modeling: Illustration
30	Knowledge-based Modeling also called ‘fold recognition’, ‘threading’, ‘recognition of remote homologies’ sequences with low % identity can adopt the same 3D fold around 1000 folds are expected to occur in all proteins sequence is ‘threaded’ into each of the known folds in a database and resulting energy is evaluated by a score Methods: e.g. FUGUE, 3D-PSSM www.sbg.bio.ic.ac.uk/~3dpssm/
31	Knowledge-based Modeling: Flow Chart
32	Ab Initio Protein Folding Rosetta method – based on local structures observed in PDB; MC search of the possible combinations of likely local structures, minimizing a scoring function that accounts for nonlocal interactions such as compactness, hydrophobic burial, and pair interactions Static approaches – search for kinetically accessible minima in hypersurfaces generated by force-field-type description of electrostatic, H-bonding and dispersion interactions Dynamic approaches – simulations with different protein representations (coarse-grain Langevin simulations, torsional angular simulations, MD simulations
33	Protein Structure Databases PDB --- www.rcsb.org/pdb Derivative databases: SumPDB – annotated PDB --- www.biochem.ucl.ac.uk/bsm/pdbsum/ Relibase+ - based on PDB, focusing on ligands and binding sites --- relibase.rutgers.edu/ MSD – Macromolecular Structure Database --- www.ebi.ac.uk/msd/
34	What are the Best Methods? CASP – Community-wide experiment on the Critical Assessment of techniques for protein Structure Prediction - http://predictioncenter.llnl.gov/ a competition organized every 2 years since 1994 experimentalists provide unpublished structures the sequences are made available to all researchers researchers can predict the structure by --- homology modeling --- fold recognition (threading) --- ab initio modeling results are then compared against the real experimental structure.
35	Quality of the Predictions I comparative modeling – CASP3
36	Quality of the Predictions II threading, ab initio – CASP3