Computational Molecular Biology (

Transcription

1 Computational Molecular Biology ( cmgm.stanford.edu/biochem218/) Biochemistry 218/Medical Information Sciences 231 Douglas L. Brutlag, Lee Kozar Jimmy Huang, Josh Silverman

2 Lecture Syllabus ( cmgm.stanford.edu/biochem218/) Doug Brutlag, 1999 Date Topic Lecturer Sept. 22 Representations of Sequences and Structures Doug Brutlag Sept 27 PubMed & Full Text Journal Access Janet Morrison Sept 29 Molecular Biology Databases on the Web Doug Brutlag Oct 4 Molecular Databases II Doug Brutlag Oct 6 Pattern Matching with Consensus Sequences Doug Brutlag Oct. 11 Quantitative and Probabilistic Pattern Matching Doug Brutlag Oct. 13 Sequence Alignment Doug Brutlag Oct. 18 Rapid Sequence Similarity Search I Doug Brutlag Oct. 20 Rapid Sequence Similarity Search II Doug Brutlag Oct. 25 Near-Optimal Sequence Alignments Doug Brutlag Oct. 27 Multiple Sequence Alignment Doug Brutlag Nov 1 Sequence Based Phylogenies Doug Brutlag Nov. 3 Sequence Blocks and Profiles Doug Brutlag Nov. 8 Discrete Protein Sequence Motifs Doug Brutlag Nov. 10 Protein Microenvironments Russ Altman Nov. 15 Probabilistic Protein Motifs Tom Wu Nov. 17 Motif Discovery Using Gibb's Sampling Scott Schmidler Nov. 22 Nov. 24 Issues in Predicting Protein Secondary Structure Scott Schmidler Protein Folds and Protein Structure SuperpositionAmit P. Singh Nov. 30 Protein Ligand Docking Amit P. Singh

3 Course Availability Gates B03 Monday and Wednesday 2:15-3:45 PM Stanford Center for Professional Development scpd.stanford.edu/ Live on SITN Channel E1 Stanford Online Course available 24 hours/day, 7 days/week Students may register in any quarter

4 Course Requirements Lectures Theoretical background of current methods Strengths and weaknesses of current approaches Future directions for improvements Demonstrations Implementations (Mac, PC, Unix, Web) Illustrate homework Six to eight homework assignments All homework submitted electronically as attachments Doug Brutlag, 1999 Due one week after assigned Final project (DUE NOVEMBER 30TH) Critically review an area Critically analyze your own data sets Propose new approach Implement a new approach

5 Bioinformatics Text

6 Reviews

7 Durbin et al.

8 Gusfield

9 Baldi Bioinformatics

10 Genomics, Bioinformatics & Medicine Genomics Molecular Diagnostics Molecular Epidemiology Bioinformatics Identify Drug Targets Rational Drug Design Genetic Therapy

11 Genomics, Bioinformatics & Medicine Genomics Molecular Diagnostics Molecular Epidemiology Bioinformatics Identify Drug Targets Rational Drug Design Genetic Therapy Machine Learning Artificial Intelligence Algorithms Robotics Statistics & Probability Graph Theory Databases Information Theory

12 National Center for Biotechnology Information (

13 Human Genome Resources ( ncbi.nlm.nih.gov/genome/guide/)

14 Genes on Chromosome 7 ( genemap/map. /map.cgi?chr=7)

15 Central Paradigm of Molecular Biology DNA RNA Protein Phenotype Molecules Structure Function Processes Mechanism Specificity Regulation

16 Central Paradigm of Bioinformatics Genetic Information SRAAINKHIVA VSYQTVSRVVN VSTATVSRALA GVTTTVSHVIN SGVSAVSAILN GVSEMTRRDLN TAYATIHVRVE GSQPTVSRELA MSIATITRGSN ISRETVGRILK FDISRLSHLFR LRPSRLAHLFR MTVETISRLLG TLEFHLHRLFK

17 Central Paradigm of Bioinformatics Genetic Information Molecular Structure SRAAINKHIVA VSYQTVSRVVN VSTATVSRALA GVTTTVSHVIN SGVSAVSAILN GVSEMTRRDLN TAYATIHVRVE GSQPTVSRELA MSIATITRGSN ISRETVGRILK FDISRLSHLFR LRPSRLAHLFR MTVETISRLLG TLEFHLHRLFK

18 Central Paradigm of Bioinformatics Genetic Information Molecular Structure Biochemical Function SRAAINKHIVA VSYQTVSRVVN VSTATVSRALA GVTTTVSHVIN SGVSAVSAILN GVSEMTRRDLN TAYATIHVRVE GSQPTVSRELA MSIATITRGSN ISRETVGRILK FDISRLSHLFR LRPSRLAHLFR MTVETISRLLG TLEFHLHRLFK

19 Central Paradigm of Bioinformatics Genetic Information Molecular Structure Biochemical Function Symptoms (Phenotype) SRAAINKHIVA VSYQTVSRVVN VSTATVSRALA GVTTTVSHVIN SGVSAVSAILN GVSEMTRRDLN TAYATIHVRVE GSQPTVSRELA MSIATITRGSN ISRETVGRILK FDISRLSHLFR LRPSRLAHLFR MTVETISRLLG TLEFHLHRLFK

20 Central Paradigm of Bioinformatics Genetic Information Molecular Structure Biochemical Function Symptoms (Phenotype) SRAAINKHIVA VSYQTVSRVVN VSTATVSRALA GVTTTVSHVIN SGVSAVSAILN GVSEMTRRDLN TAYATIHVRVE GSQPTVSRELA MSIATITRGSN ISRETVGRILK FDISRLSHLFR LRPSRLAHLFR MTVETISRLLG TLEFHLHRLFK

21 Central Paradigm of Bioinformatics Genetic Information Molecular Structure Biochemical Function Phenotype

22 Challenges Understanding Genetic Information Genetic Information Molecular Structure Biochemical Function Phenotype Genetic information is redundant Structural information is redundant Single genes have multiple functions Genes are one dimensional but function depends on three-dimensional structure

23 Redundancy in Genomic Sequences DNA is double-stranded Genetic code Acceptable amino-acid replacements Intron-exon variation Strain variation Sequencing errors

24 Multiple Representations of Sequences

25 Multiple Representations of Sequences Sequences of Common Structure or Function Doug Brutlag, 1999 Sequence Alignments VLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHF------DLSHGS : : : : : : : : : : : 2 HLTPEEKSAVTALWGKV--NVDEVGGEALGRLLVVYPWTQRFFESFGDLSTPDAVMGN Initial Score = 63 Optimized Score = 98 Significance = 5.51 Residue Identity = 14% Matches = 21 Mismatches = 22 Gaps = 2 Conservative Substitutions = 11

26 Multiple Representations of Sequences Consensus Sequences Zinc Finger (C2H2 type) CX{2,4}CX{12}HX{3,5}H Sequences of Common Structure or Function Doug Brutlag, 1999 Sequence Alignments VLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHF------DLSHGS : : : : : : : : : : : 2 HLTPEEKSAVTALWGKV--NVDEVGGEALGRLLVVYPWTQRFFESFGDLSTPDAVMGN Initial Score = 63 Optimized Score = 98 Significance = 5.51 Residue Identity = 14% Matches = 21 Mismatches = 22 Gaps = 2 Conservative Substitutions = 11

27 Multiple Representations of Sequences Blocks, Profiles or Templates Position A R N D C Q E G H I L K M F P S T W Y V Doug Brutlag, 1999 Consensus Sequences Zinc Finger (C2H2 type) CX{2,4}CX{12}HX{3,5}H Sequences of Common Structure or Function Sequence Alignments VLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHF------DLSHGS : : : : : : : : : : : 2 HLTPEEKSAVTALWGKV--NVDEVGGEALGRLLVVYPWTQRFFESFGDLSTPDAVMGN Initial Score = 63 Optimized Score = 98 Significance = 5.51 Residue Identity = 14% Matches = 21 Mismatches = 22 Gaps = 2 Conservative Substitutions = 11

28 Multiple Representations of Sequences Blocks, Profiles or Templates Position A R N D C Q E G H I L K M F P S T W Y V Doug Brutlag, 1999 Consensus Sequences Zinc Finger (C2H2 type) CX{2,4}CX{12}HX{3,5}H Sequences of Common Structure or Function Sequence Alignments Hidden Markov Model VLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHF------DLSHGS : : : : : : : : : : : 2 HLTPEEKSAVTALWGKV--NVDEVGGEALGRLLVVYPWTQRFFESFGDLSTPDAVMGN Initial Score = 63 Optimized Score = 98 Significance = 5.51 Residue Identity = 14% Matches = 21 Mismatches = 22 Gaps = 2 Conservative Substitutions = 11 AA1 D 2 D 3 D 4 D 5 I 1 I 2 I 3 I 4 I 5 AA2 AA3 AA4 AA5 AA6

29 Multiple Representations of Protein Structure K b (b i - b o )2 + all bonds K ( i - o )2 + all angles all torsion angles all non-bonded pairs all partial charge pairs AA1 K [1 - cos (n i + )] + {(r o /r ij )12-2 (r o /r ij )6} q i q j / r ij AA6 82% Hydrophobic 18% Hydrophilic AA2 AA3 HTH? AA4 AA FPTTKTYFPHF-DLS-----HGS : : : YPWTQRFFESFGDLSTPDAVMGN 40 50

30 Sequence Alignment ( alion/) X X F--SGGNTHIYMNHVEQCKEILRREPKELCELVISGLPYKFRYLSTKE-QLK-Y : :: : : : : : ::::: :: GDFIHTLGDAHIYLNHIEPLKIQLQREPRPFPKLRILRKVEKIDDFKAEDFQIEGYN X X Region End Score = Similarity-weights - Penalties Region Start where: Region End Region Start Penalty = Gap-penalty + Size-of-gap x Gap-size-penalty

31 Smith-Waterman Similarity Search Query: HU-NS1 Maximal Score: 452 PAM Matrix: 200 Gap Penalty: 5 Gap Extension: 0.5 No. Score Match Length DB ID Description Pred. No DBHB_ECOLI DNA-BINDING PROTEIN H 8.74e DBHB_SALTY DNA-BINDING PROTEIN H 1.54e DBHA_ECOLI DNA-BINDING PROTEIN H 1.64e DBHA_SALTY DNA-BINDING PROTEIN H 1.64e DBH_BACST DNA-BINDING PROTEIN I 1.35e DBH_BACSU DNA-BINDING PROTEIN I 1.35e DBH_VIBPR DNA-BINDING PROTEIN H 2.35e DBH_PSEAE DNA-BINDING PROTEIN H 2.14e DBH1_RHILE DNA-BINDING PROTEIN H 1.47e DBH_CLOPA DNA-BINDING PROTEIN H 2.52e DBH_RHIME DNA-BINDING PROTEIN H 3.18e DBH5_RHILE DNA-BINDING PROTEIN H 9.29e DBH_ANASP DNA-BINDING PROTEIN H 3.32e DBH_CRYPH DNA-BINDING PROTEIN H 2.70e DBH_THETH DNA-BINDING PROTEIN I 1.07e IHFA_SERMA INTEGRATION HOST FACT 4.46e IHFA_RHOCA INTEGRATION HOST FACT 3.52e IHFA_SALTY INTEGRATION HOST FACT 5.90e IHFA_ECOLI INTEGRATION HOST FACT 9.87e IHFB_ECOLI INTEGRATION HOST FACT 7.71e IHFB_SERMA INTEGRATION HOST FACT 7.71e TF1_BPSP1 TRANSCRIPTION FACTOR 3.42e DBH_THEAC DNA-BINDING PROTEIN H 2.12e GLGA_ECOLI GLYCOGEN SYNTHASE (EC 3.80e-01 Doug Brutlag, 1999

32 Decypher Similarity Search ( decypher.stanford.edu/)

33 Prosite Consensus Patterns ( expasy.ch/prosite/) Active site of trypsin-like serine proteases G D S G G Zinc Finger (C 2 H 2 type) C.{2,4} C.{12} H.{3,5} H N-Glycosylation Site N [^P] [S T] [^P] Homeobox Domain Signature [LIVMF].{5} [LIVM].{4} [IV] [RKQ]. W.{8} [RK]

34 The Optimal Way to Develop Patterns ( ch/images/cartoon/ /images/cartoon/prosite.gif)

35 EMOTIF Pattern Discovery (

36 Identifying Protein Functions ( emotif-search)

37 Identifying Protein Function

38 Mapping Sequence Motifs to Structural Motifs (

39 Motifs as Potential Drug Targets HIV Reverse Transcriptase o..y[vlim]dd[vli]oo.ii

40 ematrix: : Position-Specific Scoring Matrices ( Position Structural or functional motif Examples of motif HSGEQLAETLGMSRAAINKHIQ VTLYDVAEYAGVSYQTVSRVVN AMIKDVALKAKVSTATVSRALM ATIKDVAKRAGVSTTTVSHVIN ITIYDLAELSGVSASAVSAILN LHLKDAAALLGVSEMTIRRDLN TAYAELAKQFGVSPGTIHVRVE GSLTEAAHLLGTSQPTVSRELA MSQRELKNELGAGIATITRGSN ITRQEIGQIVGCSRETVGRILK FDIASVAQHVCLSPSRLSHLFR LRIDEVARHVCLSPSRLAHLFR MTRGDIGNYLGLTVETISRLLG VTLEALADQVGMSPFHLHRLFK. A R N D C Q E G H I L K M F P S T W Y V

41 ematrix-search exon/ematrix/ematrix-search.html

42 ematrix Search Results (

43 Block Signatures for a Protein Family ( fhcrc.org/) INKHIQ VSRVVN ASRALM VSHVIN VSAILN IRRDLN THVRVE GSSELA MTRGSN VGRILK LSHLFR LAHLFR ISRLLG LHRLFK HSGEQLAETLGMSRAAINKHIQ VTLYDVAEYAGVSYQTVSRVVN AMIKDVALKAKVSTATVSRALM ATIKDVAKRAGVSTTTVSHVIN ITIYDLAELSGVSASAVSAILN LHLKDAAALLGVSEMTIRRDLN TAYAELAKQFGVSPGTIHVRVE GSLTEAAHLLGTSQPTVSRELA MSQRELKNELGAGIATITRGSN ITRQEIGQIVGCSRETVGRILK FDIASVAQHVCLSPSRLSHLFR LRIDEVARHVCLSPSRLAHLFR MTRGDIGNYLGLTVETISRLLG VTLEALADQVGMSPFHLHRLFK SRAAINKHIVA VSYQTVSRVVN VSTATVSRALA GVTTTVSHVIN SGVSAVSAILN GVSEMTRRDLN TAYATIHVRVE GSQPTVSRELA MSIATITRGSN ISRETVGRILK FDISRLSHLFR LRPSRLAHLFR MTVETISRLLG TLEFHLHRLFK

44 Hidden Markov Models (after Haussler) D 2 D 3 D 4 D 5 I 1 I 2 I 3 I 4 I 5 AA1 AA2 AA3 AA4 AA5 AA6

45 Multiple Representations of Sequences Blocks, 1 2 Profiles or 7 Templates Doug Brutlag, 1999 Position A R N D C Q E G H I L K M F P S T W Y V Consensus Sequences Zinc Finger (C2H2 type) CX{2,4}CX{12}HX{3,5}H Sequences of Common Structure or Function Sequence Alignments Hidden Markov Model VLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHF------DLSHGS : : : : : : : : : : : 2 HLTPEEKSAVTALWGKV--NVDEVGGEALGRLLVVYPWTQRFFESFGDLSTPDAVMGN Initial Score = 63 Optimized Score = 98 Significance = 5.51 Residue Identity = 14% Matches = 21 Mismatches = 22 Gaps = 2 Conservative Substitutions = 11 AA1 D 2 D 3 D 4 D 5 I 1 I 2 I 3 I 4 I 5 AA2 AA3 AA4 AA5 AA6

46 Protein Identification BLAST No Smith & Waterman No MOTIFS No HMMs No Yes Yes Yes Yes Homolog Homolog Motif Superfamily PROFILES Yes No SCOP Yes No Molecular or Structural Biology Maybe Domain Domain Function

47 Sequence Representations Consensus Deterministic Alignment Blocks or Weight Matrices Templates or Profiles Bayesian Networks Hidden Markov Models Probabilistic