Applying Semantic Technologies to the Glycoproteomics Domain

1
Applying Semantic Technologies to the
Glycoproteomics Domain
W. S YorkMay 15, 2006
2
Some Goals of Glycoproteomics

• How do changes in the expression levels of
  specific genes alter the expression of specific
  glycans on the cell surface?
• Are changes in the expression of specific
  glycans at the cell surface related to cell
  function, cell development, and disease?
• What are the mechanisms by which specific
  glycans at the cell surface affect cell function,
  cell development, and the progression of disease?

3
Challenges of Glycoproteomics

• Vast amounts of data collected by
  high-throughput experiments - better methods for
  data archival, retrieval, and analysis are needed
• Complex structures of glycans and glycoproteins
  better methods for representing branched
  structures and finding structural and functional
  homologies are needed
• Complex Biology and Biochemistry better
  methods to find relationships between the
  glycoproteome and biological processes are needed

4
Glycoproteomics Solutions

• Brute-force analysis of flat data files
• Too much data
• Data is heterogeneous
• What does the data represent?
• Relational databases
• Data is well organized
• Data organization is relatively rigid
• What does the data represent?
• Semantic Technologies
• Data is well organized
• Data organization is flexible
• Concepts represented by data are accessible
• Relationships between concepts are accessible

5
What is Semantic Technology?
The implication is that enabling computers to
understand the meanings of and relationships
between concepts will allow them to reason and
communicate in a way that is analogous to the way
humans do.
Semantics1. (Linguistics) The study or science
of meaning in language.2. (Linguistics) The
study of relationships between signs and symbols
and what they represent. The American Heritage
Dictionary of the English Language, Fourth
Edition
Semantic Technology The use of formal
representations of concepts and their
relationships to enable efficient, intelligent
software.
Ontology (Computer Science) A model that
represents a domain and is used to reason about
the objects in that domain and the relations
between them. http//en.wikipedia.org/wiki/Ontolog
y_(computer_science)
6
A Simple Ontology
Organism
is_a
is_a
Animal
Plant
is_a
is_a
is_a
is_a
is_a
Lion
Deer
Cow
Hosta
Alfalfa
is_a
is_a
is_a
is_a
is_a
Elsa
Elsie
Bambi
My Hosta
Peters Alfalfa
ate
is_a
ate
ate
ate
Simba
7
A Simple Ontology
Organism
is_a
is_a
Animal
Plant
eats
is_a
eats
is_a
Carnivore
Herbivore
is_a
is_a
is_a
is_a
is_a
Lion
Deer
Cow
Hosta
Alfalfa
is_a
is_a
is_a
is_a
is_a
Elsa
Elsie
Bambi
My Hosta
Peters Alfalfa
ate
is_a
ate
ate
ate
Simba
8
The Structure of GlycO Concept Taxonomy
chemical entity
9
Concept Taxonomy
The Structure of GlycO
10
Concept Taxonomy
Instances and Properties
The Structure of GlycO
N-glycan_00020
has_residue
is_linked_to
residue
is_instance_of
glycan moiety
N-glycan core b-D-Manp
N-glycan a-D-Manp 4
N-glycan
is_a
amino acid residue
is_a
is_instance_of
is_instance_of
O-glycan
carbohydrate residue
11
The GlycO Ontology in Protégé
3 Top-Level Classes are Defined in GlycO
12
The GlycO Ontology in Protégé
Semantics Include Chemical Context
This Class Inherits from 2 Parents
13
The GlycO Ontology in Protégé
The ?-D-Manp residues in N-glycans are found in
8 different chemical environments
14
GlycoTree A Canonical Representation of
N-Glycans
We give a residue in this position the same
name, regardless of the specificstructure it
resides in
Semantics!
N. Takahashi and K. Kato, Trends in Glycosciences
and Glycotechnology, 15 235-251
15
The GlycO Ontology in Protégé
Bisecting ?-D-GlcpNAc
16
The GlycO Ontology in Protégé
17
The GlycO Ontology in Protégé
1,3-linked ??-L-Fucp
18
The GlycO Ontology in Protégé
19
Ontology Population Workflow
20
Ontology Population Workflow
Asn(41)b-D-GlcpNAc (41)b-D-GlcpNA
c (41)b-D-Manp
(31)a-D-Manp (21)b-D-GlcpNAc
(41)b-D-GlcpNAc
(61)a-D-Manp (21)b-D-GlcpNAc

21
Ontology Population Workflow
ltGlycangt ltaglycon name"Asn"/gt ltresidue
link"4" anomeric_carbon"1" anomer"b"
chirality"D" monosaccharide"GlcNAc"gt
ltresidue link"4" anomeric_carbon"1" anomer"b"
chirality"D" monosaccharide"GlcNAc"gt
ltresidue link"4" anomeric_carbon"1" anomer"b"
chirality"D" monosaccharide"Man" gt
ltresidue link"3" anomeric_carbon"1" anomer"a"
chirality"D" monosaccharide"Man" gt
ltresidue link"2" anomeric_carbon"1" anomer"b"
chirality"D" monosaccharide"GlcNAc" gt
lt/residuegt ltresidue link"4"
anomeric_carbon"1" anomer"b" chirality"D"
monosaccharide"GlcNAc" gt lt/residuegt
lt/residuegt ltresidue link"6"
anomeric_carbon"1" anomer"a" chirality"D"
monosaccharide"Man" gt ltresidue
link"2" anomeric_carbon"1" anomer"b"
chirality"D" monosaccharide"GlcNAc"gt
lt/residuegt lt/residuegt lt/residuegt
lt/residuegt lt/residuegt lt/Glycangt
22
The ProPreO Ontology in Protégé
3 Top-Level Classes are Defined in ProPreO
23
The ProPreO Ontology in Protégé
This Class Inheritsfrom 2 Parents
24
The ProPreO Ontology in Protégé
This Class Inheritsfrom 2 Parents
25
Semantic Annotation of MS Data
parent ion charge
830.9570 194.9604 2 580.2985
0.3592 688.3214 0.2526 779.4759
38.4939 784.3607 21.7736 1543.7476
1.3822 1544.7595 2.9977 1562.8113
37.4790 1660.7776 476.5043
parent ion m/z
parent ionabundance
fragment ion m/z
fragment ionabundance
ms/ms peaklist data
26
Semantically Annotated MS Data
ltms/ms_peak_listgt ltparameter instrumentmicromass
_QTOF_2_quadropole_time_of_flight_mass_spectromete
r mode ms/ms/gt ltparent_ion m/z
830.9570 abundance194.9604 z2/gt ltfragment_ion
m/z 580.2985 abundance 0.3592/gt lt
fragment_ion m/z 688.3214 abundance
0.2526/gt lt fragment_ion m/z 779.4759
abundance 38.4939/gt lt fragment_ion m/z
784.3607 abundance 21.7736/gt lt fragment_ion
m/z 1543.7476 abundance 1.3822/gt lt
fragment_ion m/z 1544.7595 abundance
2.9977/gt lt fragment_ion m/z 1562.8113
abundance 37.4790/gt lt fragment_ion m/z
1660.7776 abundance 476.5043/gt ltms/ms_peak_listgt

OntologicalConcepts
27
(No Transcript)
28
An Integrated Semantic Information System

• Formalized domain knowledge is in ontologies
• The schema defines the concepts
• Instances represent individual objects
• Relationships provide expressiveness
• Data is annotated using concepts from the
  ontologies
• The semantic annotations facilitate the
  identification and extraction of relevant
  information
• The semantic relationships allow knowledge that
  is implicit in the data to be discovered

29
Satya Sahoo Christopher Thomas Cory Henson Ravi
Pavagada
Amit Sheth Krzysztof Kochut John Miller
James Atwood Lin Lin Alison Nairn Gerardo
Alvarez-Manilla Saeed Roushanzamir
Michael Pierce Ron Orlando Kelley
Moremen Parastoo Azadi
Alfred Merrill