Characterize the stoichiometric reactions of a metabolic model is a
required and time-consuming work. The MINVAL package
includes the characterizeReactions
function to characterize
the stoichiometric functions and metabolites by compartment.
To load the MINVAL package just type:
library(minval)
To show the potential use of the characterizeReactions
function of the MINVAL package the Human Metabolic
Reconstruction (RECON 2.04) was included in a human-readable format. To
load it just type:
<- read.csv(system.file("extdata", "rRECON2.csv",
RECON package = "minval"))
dim(RECON)
## [1] 7441 8
The characterizeReactions
function requires a set of
stoichiometric reactions as input. The given stoichiometric reactions
must have the following mandatory characteristics:
'=>'
or
'<=>'
(Inverse arrow symbols '<='
or
other types as: '-->'
, '<==>'
,
'->'
will not be parsed and will lead to errors.)+
) must be surrounded by
a space character.-
) symbol.[compartment]
) joined at the end of metabolite name.The characterizeReactions
function: - Counts the number
of reactions - Computes the relative frequency of each reaction type
(transport, exchange and compartmentalized) - Computes the relative
frequency of reactions by compartment - Counts the number of unique
metabolites - Computes the relative frequency of metabolites by
compartment
Finally the characterizeReactions
function returns all
these information as a labeled list. To characterize a set of
stoichiometric reactions just type:
<- characterizeReactions(reactionList = RECON$REACTION)
RECONcharacteristics
RECONcharacteristics## $nReactions
## [1] 7441
##
## $rType
##
## Compartmentalized reaction Exchange reaction
## 55.825830 9.420777
## Transport reaction
## 34.753393
##
## $cReaction
##
## c e g l m n r x
## 24.593469 1.760516 3.628545 2.983470 9.958339 1.666443 6.208843 5.026206
##
## $nMetabolites
## [1] 5063
##
## $cMetabolites
##
## c e g l m n r x
## 37.092633 12.680229 6.261110 5.964843 14.892356 3.258937 11.258147 8.591744
Computed values can be easy plotted as follows:
Metabolic models generally includes three types of reactions:
h[m] + nadph[m] + o2[m] + 25hvitd2[m] => h2o[m] + nadp[m] + 1a25dhvitd2[m]
2 hco3[e] + na1[e] <=> 2 hco3[c] + na1[c]
acetone[e] <=>
To plot it just type:
pie(x = RECONcharacteristics$rType,
main = "Reactions by Type"
)
Compartmentalized reactions identified also are categorized and reported as a relative frequency to their associated compartment. To plot it just type:
pie(x = RECONcharacteristics$cReaction,
main = "Reactions by Compartment",
labels = compartmentNames
)
Metabolites are categorized and reported as a relative frequency to their associated compartment. To plot it just type:
pie(x = RECONcharacteristics$cMetabolites,
main = "Metabolites by Compartment",
labels = compartmentNames
)