Download bchm628_lect5_15

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts

List of types of proteins wikipedia, lookup

Protein structure prediction wikipedia, lookup

Cyclol wikipedia, lookup

Protein wikipedia, lookup

Protein–protein interaction wikipedia, lookup

Proteomics wikipedia, lookup

Intrinsically disordered proteins wikipedia, lookup

Circular dichroism wikipedia, lookup

Nuclear magnetic resonance spectroscopy of proteins wikipedia, lookup

Protein moonlighting wikipedia, lookup

Western blot wikipedia, lookup

G protein–coupled receptor wikipedia, lookup

Bimolecular fluorescence complementation wikipedia, lookup

Protein purification wikipedia, lookup

Trimeric autotransporter adhesin wikipedia, lookup

Protein mass spectrometry wikipedia, lookup

P-type ATPase wikipedia, lookup

Homology modeling wikipedia, lookup

Structural alignment wikipedia, lookup

Protein folding wikipedia, lookup

Protein domain wikipedia, lookup

Protein design wikipedia, lookup

Rosetta@home wikipedia, lookup

Transcript
Predicting Function
(& location & post-tln modifications)
from Protein Sequences
June 15, 2015
Outline
 Usefulness of protein domain analysis
 Types of protein domain databases
 Interpro scan of multiple domain DB
 Using the SMART database
 Predicting post-translational modifications
When annotation is NOT enough
 You’ve got a list of genes, most of which have
been annotated with gene ontology and a
potential protein function
 Why would you want to go on and look more
specifically at the protein domains?
Limitations of annotation
 Even in a model organism with large amount of
resources, most genes are still annotated by
similarity
 Often, the name given is based on the BEST match
to a particular domain or known protein
 But…
Limitations of BLAST
 Likelihood of finding a homolog to a sequence:
 >80% bacteria
 >70% yeast
 ~60% animal
 Rest are truly novel sequences
 ~900/6500 proteins in yeast without a known
function
 NAME: Similar to yeast protein YAL7400 not very
informative
Limitations of similarity
 Proteins with more than one domain cause
problems.
 Numerous matches to one domain can mask
matches to other domains.
 Increased size of protein databases
 Number related sequences rises and less related
sequence hits may be lost
 Low-complexity regions can mask domain
matches
Proteins are modular
 Individual domains can and often do fold
independently of other domains within the same
protein
 Domains can function as an independent unit (or
truncation experiments would never work)
 Thus identity of ALL protein domains within a
sequence can provide further clues about their
function
Proteins can have >1 domain
The name: protein kinase receptor UFO doesn’t
necessarily tell you that this protein also contains IgG and
fibronectin domains or that it has a transmembrane
domain
Domains are not always functional
 If a critical residue is
missing in an active site, it’s
not likely to be functional
 A similarity score won’t
pick that up
Multiple protein domain databases
Protein signature databases
 Identify domains or classify proteins into families to
allow inference of function
 Approaches include:
 regular expressions and profiles
 position-specific scoring matrix-based fingerprints
 automated sequence clustering
 Hidden Markov Models (HMMs)
PROSITE
 Regular expression patterns describing functional
motifs
M-x-G-x(3)-[IV]2-x(2)-{FWY}
 Enzyme catalytic
sites
 Prosthetic group attachment sites
 Ligand or metal binding sites
 Either matches or not
 Some families/domains defined by co-occurrence
Citrate synthase
G-[FYAV]-[GA]-H-x-[IV]-x(1,2)-[RKTQ]-x(2)-[DV]-[PS]-R
PRINTS
 Similar to PROSITE patterns
 Multiple-motif approach using either identity or
weight-matrix as basis
 Groups of conserved motif provide diagnostic
protein family signatures
 Can be created at super-family, family and sub-
family level
http://www.bioinf.manchester.ac.uk/dbbrowser/PRINTS/index.php
Profile-HMMs
 Models generated from alignments of many homologues then
counting frequency of occurrence for each amino acid in each
column of the alignment (profile).
 Profile-HMMs used to create probabilities of occurrence
against background evolutionary model that accounts for
possible substitutions.
 Provides convenient and powerful way of identifying
homology between sequences.
 Find domains in sequences that would never be found by
BLAST alone
HMM domain databases
 Pfam

Classify novel sequences into protein domain profiles

Most comprehensive; >13,000 protein families (v26)
 SMART

Signaling, extracellular and chromatin proteins

Identification of catalytic site conservation for enzymes
 TIGRFAMs

Families of proteins from prokaryotes
 PANTHER

Classification based on function using literature evidence
PFAM
 >16,230 manually curated profiles
 Can use the profile to search a genome for
matches
Can submit a protein to PFAM
 Limited to single protein submission
 Output gives you an e-value that estimates the
likelihood that the domain is there
 Up to you to determine if domain is functional
http://pfam.xfam.org
Keyword search
PFAM Summary
PFAM Domain Organization
PFAM Interactions
SMART database
 SMART: Simple Modular Architecture Research Tool
 Focus on signaling, extracellular and chromatin-
associated proteins
 Curated models for >1200 domains
 Use?
 I have several kinase domains in my protein list and
want to know which ones are functional.
 What other domains are found in signaling proteins?
Search for matches
Uniprot or Ensemble
Protein Accession number
Protein sequence
Add other searches
SMART Output
Mouse over for
information
Prediction of FUNCTIONAL catalytic activity
Can browse the domains
InterPro Scan
 Combines search methods from several protein
databases
 Uses tools provided by member databases
 Uses threshold scores for profiles & motifs
 Interpro convenient means of deriving a
consensus among signature methods
Define which
domain databases
to search
Example InterProScan search
 Submitting an olfactory receptor gene (member
of the GPCR class of proteins) to InterPro
InterPro
family
2nd
InterPro
family
 Submitting a different human GPCR protein to
Interpro
Same
InterPro
family
New
InterPro
family
InterProScan Families
InterProScan annotation
SMART & PFAM search
SMART DB results:
PFAM DB results:
Are 2 proteins homologs?
 S. cerevisiae Ste3 is a GPCR pheromone receptor
 Similarity to C. gatti protein:

25% identical, 45% similar, E-value 10-25
Very similar domain content and
arrangement
Advantage of InterProScan
 Interpro integrates the different databases to
create a protein family signature.
 Pfam/SMART/PANTHER/Gene3D & TIGR-FAM
will find domain families
 PROSITE can find very specific signature patterns
 PRINTS can distinguish related members of same
protein family
Cannot change the statistical cut-off for what is considered a
significant match
Function from sequence
 Membrane bound or secreted?
 GPI anchored?
 Cellular localization?
 Post-translational modification sites?
CBS prediction services
 Protein sorting
 SignalP, TargetP, others
 Post-translational modification
 Acetylation, phosphorylation, glycosylation
 Immunological features
 Epitopes, MHC allele binding, ect
 Protein function & structure
 Transmembrane domains, co-evolving positions
Transmembrane domain prediction
Phosphorylation prediction
O-glycosylation
EMBOSS
Open source software for molecular biology
 Predict antigenic sites
 Useful if want to design a peptide antibody
 Look for specific motifs, even degenerate
 Known phosphorylation motifs
 Find motifs in multiple sequences with one
submission
 Get stats on proteins/nucleic acid sequences
 Sequence manipulation of all kinds
Today in lab
 Tutorial on protein information sites
 From a sublist generated using DAVID, generate a
list of protein IDs and obtain the sequences
 Obtain protein accession numbers for the cluster
 Submit to SMART database to
characterize/analyze the domains
 Pick 2 proteins to do additional predictions