SYSTEMS BIOLOGY
Course 2024-2025
Raúl Guantes, David Míguez
raul.guantes@uam.es
Facultad de Ciencias, Módulo 08 401-12
http://sysbio.openwetware.org
Good textbooks that cover almost all modeling approaches to systems
biology:
- Uri Alon. Introduction to systems biology (CRC, 2020 New edition). A very
personal view of biological systems (that I follow in part in this course).
- R. Phillips, J. Kondev, J. Theriot, H. G. García, Physical Biology of the Cell,
Garland Science 2013 (2nd edition). Contains many ideas and concepts from
physics.
- E. Klipp, Systems Biology: a text book. Princeton University Press, 2016
(second edition). A compendium of mathematical and computational methods to
model and analyze biological systems.
- E. O. Voit, A first course in Systems Biology, Garland Science, 2013. More
applied and focused on specific topics.
1. Basic concepts
1a. Numbers and quantitative estimates in Biology. Physical consequences.
1b. Dynamics in Biology: the use of ordinary and partial differential equations.
1c. Linear dynamical processes in Biology: cell population growth, protein dilution
and active degradation, constant protein production. Genome-wide quantification of
RNA and protein transcription, translation and half-lives.
1d. The response time of a non-regulated gene.
1a. The importance of quantifying and estimating scales in Biology
"
"
"
Cell Biology by the Numbers (Ron Milo and Rob Philips).
" book.bionumbers.org
" " to report a value with appropriate number of significant digits
how
" !
% Estimations and back-of-the-
%
converting between Daltons and grams !
!
you calculated 5324.1,
should it be reported as
5000, 5300, 5324.1 ?
envelope calculations are
–
+
H atom !
!
important to understand results
mH = 1 Da
!
value has
1g uncertainty NO you need to use your experience
and design experiments.
1 g of hydrogen = NA x mH mH =
6 x 1023
1.6 x 10–24g
! estimate?
Avogadro number hydrogen atom mass !
Estimation 1. You find a protein in ! what is a reasonable
uncertainty estimate?
! YES
an E. coli cell culture in a
Figure 1: Conversion between Daltons and grams. In moving from the !
concentration of 1 picoM per cell.
world of macroscopic quantities of some substance to the microscopic !
world this conversion through the Avogadro constant is often required.
say the error 10%, 2 fold, >3 fold,
! estimate is 437.21 i.e. 532 i.e. 5300 i.e. order of
Is this protein significantly
" ! magnitude only
expressed?. !
rewrite uncertainty estimate with one significant digit
!
Data: Suppose that bacteria have !
a cylindrical shape of 2 microns ! 437.12 → ±400 532 → 500 5300 → 5000 10,000
length and 1 micron diameter RULE: the last digit of the reported value (rightmost non-zero) should be located
in the same place as the one digit uncertainty value
5300 ± 400 two significant one significant order of magnitude
nanomolar in E. coli units
digits, 5300 digit, 5000 only 10,000
cell volume 1µm3 = 1 fL we prefer the convention where a leading 1 does not count as a significant digit, i.e. 1234 with one significant digit is 1200;
for order of magnitude round in log space, i.e. 300 → 100; 400 → 1000
single
molecule
c=
1 molecule
x
1 mole
x
1 fL 1
= x 10–8 M 1.6 nM Snapshot: keyto numbers
Figure 3: A flow chart help determine how intoBiology
an appropriate number of significant digits"
report values(moodle,
with
1 fL 6 x 1023 molecules 10–15 L 6
folder /materials/Cell snapshots/’
rule of thumb: 1 molecule per bacterial volume 1 nM
Figure 2: Rule of thumb for converting concentrations into absolute
The molecular ‘census’ of E. coli
Numbers have consequences in Biology
Many molecules that take part in gene expression (including DNA, transcription factors
and the enzyme polymerase) may have low copy numbers. As a consequence, gene
expression is a random process and populations of genetically identical cells have variable
numbers of each protein type.
Raser and O’Shea, Noise in gene expression: origins,
consequences, and control, Science (2005). There are many examples of
phenotypic consequences of this
randomness in gene expression
Competence/sporulation
in B. subtilis
Süel et al., Tunability and noise dependence in
differentiation dynamics, Science (2007).
Another example: average distance between proteins in a cell
d @ c -1/ 3 Estimate total number of proteins in a cell
(Nprot) and divide by cell volume
Average
distance Molecular total protein mass per cell
between ‘concentration’ (number N prot ≅
of molecules per average mass of a protein
molecules
volume)
Cell mass ~ 1 pg, ~30% is ‘dry’ weight, ~50% of ‘dry’ weight is made of proteins
total protein mass per cell ≈ 15x10 −14 gr
Average protein length ~300 aminoacids
Average aminoacid mass ~ 100 Dalton
average mass of a protein ≈ 5x10 −20 gr
total protein mass per cell 15 ×10 −14
N prot ≅ ≈ −20
=3×10 6 proteins
average mass of a protein 5 ×10
3×10 6 3
c≈ 9
molecules/nm ⇒ d ≈ 7nm
10
Typical protein radius ~ 2-5 nm
Distances between molecules are of the order of the size of the molecules!
Molecular crowding This again has physical consequences: many biochemical
reactions inside cells are limited by the slow diffusion of the
involved molecules.
1b. Dynamics in Biology
All biological systems change
and evolve in time with a huge
diversity of time scales
Dynamics in Biology: There are many examples in Biology where a transient
behavior is important. Cells sense signals that can be coding information in
amplitude, duration and frequency.
The temporal change of a single factor Global genetic programs are executed
determines cellular fate in a precise temporal ordering
Santos et al., Nature Cell Biology (2007) 324-330. Amit et al., Nature Genetics (2007) 39, 503-512.
Dynamical systems (systems that change in time) are described by differential equations.
Ordinary differential
equations (ODEs)
Linear and non-linear systems:
du
= f (u, t; µ ) If all dynamical variables u=(u1,u2,…) appear at
most with exponent 1, the system is linear.
dt Otherwise it is non-linear.
u à Dynamical variables
µ à Parameters ODEs represent homogeneous systems that
change continuously in time (deterministic)
u(t) à Trajectories (solutions)
Specify initial conditions!
Amount of protein
inside a cell (no matter
its location) as a
function of time
In many situations ODE models correspond to experiments where you get population averages
There are other systems where the spatial coordinates of the variables have to be taken
into account. An important example is in the development of animal bodies.
French flag model of a morphogen
Partial differential equations gradient interpretation
(PDEs)
¶u(x, t; µ )
= f̂ [u(x, t; µ )]
¶t
u(x,t;µ)à dynamical variables
that depend both on spatial
coordinates and time
¶
f̂ º f (u(x, t ), x, t , ; µ )
¶x
Antero-posterior position
Operator
1c. Simple dynamical processes in cells: cell growth, protein dilution and degradation
Cell growth by constant division time
Variable: N(t) (number of cells at a given time).
Question*: How many cells do I have at a given time, if cells divide after a fixed
period?
1.- Deduce by counting
Generation 0 Generation 1 Generation 2
N(Tdiv ) = 2 ⋅ N(0)
... N(2Tdiv ) = 4 ⋅ N(0)
.
Tdiv 2Tdiv .
. t
N(t) = N(0)⋅ 2 Tdiv
Passing to continuous time
2.- Which is the differential equation obeyed by N(t)?
Taking dN(t) d2 t/Tdiv t/Tdiv Ln ( 2 )
derivatives: = N(0)⋅ = N(0)⋅ 2 ⋅ = r ⋅ N(t)
dt dt Tdiv
where r=Ln(2)/Tdiv is the proliferation or growth rate (a parameter)
Simple bacterial growth follows a linear equation, which implies
exponential growth in time (solve the equation by separating variables)
N (t ) = N (0) × e rt
