Transcript Inferring genetic regulatory logic from expression data

Inferring genetic regulatory
logic from expression data
Prepared by Chunhui CAI
Logic gates

Working principles of cis-regulatory
elements




‘OR’ logic
‘AND’ logic
‘NOR’ logic
‘OR-NOR’ logic
Bayesian network

A probabilistic model

Use probability as a means to express
uncertainty about modeling variables
and their dependencies.
System and methods

The model of gene
regulatory interactions




Xi ~ regulator
Y ~ regulatee
Ii ~ intermediate
variable
θi ~ binding probability
System and methods
System and methods
System and methods

‘OR’ model
System and methods

‘OR-NOR’ model
System and methods

Rules

Gene expression values ~ two states
0 – not expressed
 1 – expressed


Type of regulation
Simultaneous
 Time delay

Result

Regulatory interactions
of 20 genes of
S.cerevisiae. The full
arcs represent
activatory regulation,
the dashed arcs
represent inhibitory
regulation. The
relationship between
genes regulating one
common gene is
described by ‘OR’function.
Discussion




Probabilistic or deterministic
Either ~ Fitting of the expression data
into state value (simply two states(0,1)?)
Or ~ Determining the
activation/repression logic
(Later) How good is our clustering
method (need to improve and more
expression data should be involved)