From biology to math and back

1
From biology to math and back

• Annotation of models

2
System biology modelling gap

• Biologist can be divided into two classes
  experimentalists who observed things that cannot
  be explained, and theoreticians who explain
  things that cannot be observed.
• Katzir-Katchalsky

3
A Circuit of Systems Biology
(Massive) data
Models
Data analysis
Software
New experiments
Predictions
??
Understanding
Model analysis
4
Top-down Bottom-up

• Top-down
• Based upon genome information
• Focused on protein function
• Generic
• Wide and shallow
• Bottom-up
• Based upon experiment and literature
• Focused on metabolites
• Very specific
• Deep and narrow

5
Hybrid approach

• Top-Down step
• Start with review paper
• Create list of key players in the pathway
• Bottom-up step for each key player
• Collect all known information about
• Activation
• Inhibition
• Synthesis
• Degradation
• Posttranslational modifications
• Interactions

6
Hybrid approach

• Top-down approach (second turn)
• Check for known cross-talks
• Repeat procedure for new list of pathways
• Especial attention
• Degradation
• Deactivation
• Switch off

7
Model reuse

• Should you repeat these steps every time
• Unfortunately yes
• Why?
• Lack of biological annotation
• Irreproducibility
• Different hypothesis
• BioModels.net
• 200 models curated(3 years)

8
MIRIAM

• Minimal Information Requested In the Annotation
  of biochemical Models
• A set of guidelines suitable for use with any
  structured format (SBML, CellML, GENESIS, CellML,
  ...) for computational models.
• MIRIAM URIs is a way to identify, in a stable and
  perennial way, elements, like proteins, genes,
  publications, etc.
• A data-type is a unique, controlled description
  of the type of the data.
• For example, if the entity is a protein, the
  data-type will be Uniprot, if the entity is an
  enzyme, the data-type will be Enzyme Nomenclature

9
MIRIAM

• MIRIAM Standard is composed of three parts
• reference correspondence,
• The model must comply with the standard in which
  it is encoded.
• The model must be clearly related to a single
  reference description.
• The encoded model structure must reflect the
  biological processes listed in the reference
  description.
• attribution annotation,
• The model has to be named.
• A citation of the reference description must be
  joined
• The name and contact of model creators must be
  joined.
• The date and time of creation and last
  modification should be specified. An history is
  useful but not required.
• external resource annotation.
• The annotation must permit to unambiguously
  relate a piece of knowledge to a model
  constituent.
• The referenced information should be described
  using a triplet data-type, identifier,
  qualifier

10
MIRIAM URI

• The usage of MIRIAM URIs permits
• uniquely describe entities with a perennial and
  stable identifier
• link an entity to one or more online resources,
  where extra knowledge about it is available
• MIRIAM URIs are not dependant on physical
  addresses (like the URLs used with a web
  browser), that is why they are persistent.
• http//www.ebi.ac.uk/compneur-srv/miriam-main/mdb?
  sectionbrowse

11
A Circuit of Systems Biology
(Massive) data
Models
Data analysis
Software
New experiments
Predictions
??
Understanding
Model analysis