Protein Fold recognition

1
Protein Fold recognition

• Morten Nielsen,
• CBS, BioCentrum,
• DTU

2
Outline

• Why model protein structure
• Classification of protein structures
• Fold, Superfamily, Family
• Protein homology modeling
• Template (fold) recognition
• Alignment
• Side chain modeling
• Loop modeling

• Reliability measures
• id bad, P-value good
• Historical overview
• Blast (simple alignment)
• Psi Blast (profiles)
• Profile-profile alignment
• Structural features
• Recombinant or democratic homology modeling
• Best methods

3
Why protein modeling?

• Experimental effort to determine protein
  structure is very large and costly
  
• The gap between the size of the protein sequence
  data and protein structure data is large and
  increasing
  
• Close to 50 of all new sequences can be homology
  modeled

4
Swiss-Prot database
5
PDB New Fold Growth
Old folds
New PDB structures
New folds

• The number of unique folds in nature is fairly
  small (possibly a few thousands)
  
• 90 of new structures submitted to PDB in the
  past three years have similar structural folds in
  PDB

6
Protein classification

• Number of protein sequences grow exponentially
• Number of solved structures grow exponentially
• Number of new folds identified very small (and
  close to constant)
  
• Protein classification can
• Generate overview of structure types
• Detect similarities (evolutionary relationships)
  between protein sequences

7
Protein structure classification
8
Classification schemes

• SCOP
• Manual classification (A. Murzin)
• CATH
• Semi manual classification (C. Orengo)
• FSSP
• Automatic classification (L. Holm)

9
Levels in SCOP

• Class Folds Superfamilies Families
• All alpha proteins 202 342 550
• All beta proteins 141 280 529
• Alpha and beta proteins (a/b) 130 213 593
• Alpha and beta proteins (ab) 260 386 650
• Multi-domain proteins 40 40 55
• Membrane and cell surface
• proteins 42 82 91
• Small proteins 72 104 162
• Total 887 1447 2630

http//scop.berkeley.edu/count.htmlscop-1.67
10
Major classes in SCOP

• Classes
• All alpha proteins
• Alpha and beta proteins (a/b)
• Alpha and beta proteins (ab)
• Multi-domain proteins
• Membrane and cell surface proteins
• Small proteins

11
All a Hemoglobin (1bab)
12
All b Immunoglobulin (8fab)
13
a/b Triosephosphate isomerase (1hti)
14
ab Lysozyme (1jsf)
15
Families

• Proteins whose evolutionarily relationship is
  readily recognizable from the sequence (25
  sequence identity)
  
• Families are further subdivided in to Proteins
• Proteins are divided into Species
• The same protein may be found in several species

16
Superfamilies

• Proteins which are (remote) evolutionarily
  related
  
• Sequence similarity low
• Share function
• Share special structural features
• Relationships between members of a superfamily
  may not be readily recognizable from the sequence
  alone

17
Folds

• Proteins which have 50 of their secondary
  structure elements arranged the in the same order
  in the protein chain and in three dimensions are
  classified as having the same fold
• No evolutionary relation between proteins
• confusingly also called fold classes

18
Links

• PDB (protein structure database)
• www.rcsb.org/pdb/
• SCOP (protein classification database)
• scop.berkeley.edu
• CATH (protein classification database)
• www.biochem.ucl.ac.uk/bsm/cath
• FSSP (protein classification database)
• www.ebi.ac.uk/dali/fssp/fssp.html

19
Superfamilies

• Proteins which are (remote) evolutionarily
  related
  
• Sequence similarity low
• Share function
• Share special structural features
• Relationships between members of a superfamily
  may not be readily recognizable from the sequence
  alone

Fold
Superfamily
Family
Proteins
20
Model accuracy. Swiss-model. 1200 models sharing
25-95 sequence identity with the submitted
sequences (www.expasy.ch/swissmod)
21
Identification of fold

• If sequence similarity is high proteins share
  structure (Safe zone)
  
• If sequence similarity is low proteins may share
  structure (Twilight zone)
  
• Most proteins do not have a high sequence
  homologous partner

Rajesh Nair Burkhard Rost Protein Science,
2002, 11, 2836-47
22
Identification of correct fold

• ID is a poor measure
• Many evolutionary related proteins share low
  sequence homology
  
• Alignment score even worse
• Many sequence will score high against every thing
  (hydrophobic stretches)
  
• P-value or E-value more reliable

23
P and E values
Score 150 10 hits with higher score (E10) 10000
hits in database P10/10000 0.001

• E-value
• Number of expected hits in database with score
  higher than match
  
• Depends on database size
• P-value
• Probability that a random hit will have score
  higher than match
  
• Database size independent

P(Score)
Score
24
Protein Homology modeling

• Identify fold (template) for modeling
• Find the structure in the PDB database that
  resembles the unknown structure the most
  
• Can be used to predict function

• Align protein sequence to template
• Simple alignment methods
• Sequence profiles
• Threading methods
• Pseudo force fields
• Model side chains and loops

25
Template identification

• Simple sequence based methods
• Align (BLAST) sequence against sequence of
  proteins with known structure (PDB database)
  
• Sequence profile based methods
• Align sequence profile (Psi-BLAST) against
  sequence of proteins with known structure (PDB)
  
• Align sequence profile against profile of
  proteins with known structure (FFAS)
  
• Sequence and structure based methods
• Align profile and predicted secondary structure
  against proteins with known structure (3D-PSSM)

26
Template identification

• Threading methods
• Align sequence against structural environment of
  proteins with known structure
  
• Use biological information
• Functional annotation in databases
• Active sites

27
Sequence profiles

• In conventional alignment, a scoring matrix
  (BLOSUM62) gives the score for matching two amino
  acids
  
• In reality not all positions in a protein are
  equally likely to mutate
  
• Some amino acids (active cites) are highly
  conserved, and the score for mismatch must be
  very high
  
• Other amino acids are mutate almost for free, and
  the score for mismatch is lower than the BLOSUM
  score
  
• Sequence profiles can capture this

28
Sequence profiles
ADDGSLAFVPSEF--SISPGEKIVFKNNAGFPHNIVFDEDSIPSGVDASK
ISMSEEDLLN TVNGAI--PGPLIAERLKEGQNVRVTNTLDEDTSIHW
HGLLVPFGMDGVPGVSFPG---I -TSMAPAFGVQEFYRTVKQGDEVTV
TIT-----NIDQIED-VSHGFVVVNHGVSME---I
IE--KMKYLTPEVFYTIKAGETVYWVNGEVMPHNVAFKKGIV--GEDAFR
GEMMTKD--- -TSVAPSFSQPSF-LTVKEGDEVTVIVTNLDE------
IDDLTHGFTMGNHGVAME---V ASAETMVFEPDFLVLEIGPGDRVRFV
PTHK-SHNAATIDGMVPEGVEGFKSRINDE----
TVNGQ--FPGPRLAGVAREGDQVLVKVVNHVAENITIHWHGVQLGTGWAD
GPAYVTQCPI
TKAVVLTFNTSVEICLVMQGTSIV----AAESHPLHLHGFNFPSNFNLVD
PMERNTAGVP
Matching any thing but G large negative score
Any thing can match
29
Sequence profiles

• Align (BLAST) sequence against large sequence
  database (Swiss-Prot)
  
• Select significant alignments and make profile
  (weight matrix) using techniques for sequence
  weighting and pseudo counts (see lecture on
  HMMs)
• Use weight matrix to align against sequence
  database to find new significant hits
  
• Repeat 2 and 3 until stop criteria

30
PDB-BLAST

• Procedure
• Build sequence profile by iterative PSI-BLAST
  search against a sequence database
  
• Use profile to search database of proteins with
  known structure (PDB)

31
Transitive BLAST

• Procedure
• Find homologues to query (your) sequence
• Find homologues to these homologues
• Etc.
• Can be implemented with e.g. BLAST or PSI-BLAST
• Also known as Intermediate Sequence Search (ISS)

32
ExampleSequence profiles

• Alignment of protein sequences 1PLC._ and 1GYC.A
• E-value 1000
• Profile alignment
• Align 1PLC._ against Swiss-prot
• Make position specific weight matrix from
  alignment
  
• Use this matrix to align 1PLC._ against 1GYC.A
• E-value

33
Sequence profiles
Score 97.1 bits (241), Expect 9e-22
Identities 13/107 (12), Positives 27/107
(25), Gaps 17/107 (15) 1PLC._ 3 ADDGSLA
FVPSEFSISPGEKI------VFKNNAGFPHNIVFDEDSIPSGVDASKIS
56 F G N
G 1GYC.A 26 ------VFPSPLITGKK
GDRFQLNVVDTLTNHTMLKSTSIHWHGFFQAGTNWADGP 79
1PLC._ 57 MSEEDLLNAKGETFEVAL---SNKGEYSFYCSP--
HQGAGMVGKVTV 98 A G F
G G G V 1GYC.A 80 AFVNQCPIAS
GHSFLYDFHVPDQAGTFWYHSHLSTQYCDGLRGPFVV 126
Rmsd3.3 Å Model red Template blue
34
Including structure

• Sequence with in a protein superfamily share
  remote sequence homology
  
• , but they share high structural homology
• Structure is known for template
• Predict structural properties for query
• Secondary structure
• Surface exposure

35
Using structure

• Sequencestructure profile-profile based
  alignments
  
• Template profiles
• Multiple structure alignments
• Sequence based profiles
• Query profile
• Sequence based profile
• Predicted secondary structure
• Position specific gap penalties derived from
  secondary structure

36
Structure biased alignment (3D-PSSM)
http//www.sbg.bio.ic.ac.uk/3dpssm/
37
Threading
Alignment score from structural fitness (pair
potential) How well does K fit environment at P6?
If P8 is acidic then fine, if P8 is basic th
en poor
Deletions
7
4
6
2
5
8
10
9
3
1
.. A T N L Y K E T L ..
Insertion
38
Threading

• Threading does not work
• The average protein does not exist
• Threading can be used in combination with
  sequence profiles, local structural features to
  improve alignment

39
CASP

• CASP
• Critical Assessment of Structure Predictions
• Every second year
• Sequences from about-to-be-solved-structures are
  given to groups who submit their predictions
  before the structure is published
  
• Modelers make prediction
• Meeting in December where correct answers are
  revealed

40
CASP5 overview
41
Successful fold recognition groups at CASP5

• 3D-Jury (Leszek Rychlewski)
• 3D-CAM (Krzysztof Ginalski)
• Template recombination (Paul Bates)
• HMAP (Barry Honig)
• PROSPECT (Ying Xu)
• ATOME (Gilles Labesse)

42
Democratic homology modeling

• Let the silent majority rule
• The highest score hit will often be wrong
• Many prediction methods will have the correct
  fold among the top 10-20 hits
  
• If many different prediction methods all have
  some fold among the top hits, this fold is
  probably correct

43
3D-Jury (Rychlewski)

• Inspired by Ab initio modeling methods
• Average of frequently obtained low energy
  structures is often closer to the native
  structure than the lowest energy structure
  
• Find most abundant high scoring model in a list
  of prediction from several predictors
  
• Use output from a set of servers
• Superimpose all pairs of structures
• Similarity score Sij of Ca pairs within 3.5Å
  (if 40else Sij0)
  
• 3D-Jury score SijSij/(N1)
• Similar methods developed by A Elofsson (Pcons)
  and D Fischer (3D shotgun)

44
LiveBench

• The Live Bench Project is a continuous
  benchmarking program. Every week sequences of
  newly released PDB proteins are being submitted
  to participating fold recognition servers. The
  results are collected and continuous evaluated
  using automated model assessment programs. A
  summary of the results is produced after several
  months of data collection. The servers must delay
  the updating of their structural template
  libraries by one week to participate

45
Meta prediction server

• Web interface to a list of public protein
  structure prediction servers
  
• Submit query sequence to all selected servers in
  one go
  
• http//bioinfo.pl/meta/

46
Meta Server
47
Meta Server
48
3D Jury
49
188 targets in total Threshold for 5 false positi
ves 50 for 3D Jury
50
Links to fold recognition servers

• Databases of links
• http//bioinfo.pl/meta/servers.html
• http//mmtsb.scripps.edu/cgi-bin/renderrelres?prot
  model
  
• Meta server
• http//bioinfo.pl/meta/
• 3DPSSM good graphical output
• http//www.sbg.bio.ic.ac.uk/servers/3dpssm/
• GenTHREADER
• http//bioinf.cs.ucl.ac.uk/psipred/
• FUGUE2
• http//www-cryst.bioc.cam.ac.uk/fugue/prfsearch.h
  tml
  
• SAM
• http//www.cse.ucsc.edu/research/compbio/HMM-apps/
  T99-query.html
  
• FOLD
• http//fold.doe-mbi.ucla.edu/
• FFAS/PDBBLAST
• http//bioinformatics.burnham-inst.org/

51
From fold to structure

• Flying to the moon has not made man conquer
  space
  
• Finding the right fold does not allow you to make
  accurate protein models
  
• Can allow prediction of protein function
• Alignment is still a very hard problem
• Most protein interactions are determined by the
  loops, and they are the least conserved parts of
  a protein structure

52
Ab initio protein modeling Modeling of newfold pr
oteins

• Only when every thing else fails
• Challenge
• Close to impossible to model Natures folding
  potential
  
• Example

53
Challenge. Folding potential

• New folds are in general constructed from a set
  of subunits, where each subunit is a part of a
  known fold.
  
• The subunits are small compared to the overall
  fold of the protein. No objective function exists
  to guide the global packing of the subunits.
  

Objective function
sij 120aa
dij 6Å
54
A way to solution

• Glue structure piece wise from fragments.
• Guide process by empirical potential (Potential
  of mean force)

Fragments with correct local structure
Natures potential
Empirical potential
55
Examples (Rosetta web server) www.bioinfo.rpi.edu/
bystrc/hmmstr/server.php
Rosetta prediction
Homology modeling
56
Take home message

• Identifying the correct fold is only a small step
  towards successful homology modeling
  
• Do not trust ID or alignment score to identify
  the fold. Use p-values
  
• Use sequence profiles and local protein structure
  to align sequences
  
• Do not trust one single prediction method, use
  consensus methods (3D Jury)
  
• Only if every things fail, use ab initio methods