Met­Ex­plore

Find­ing your path through net­work complexity

  1. OVERVIEW
  2. METABOLOME MAP­PING
  3. META­BOLIC GRAPH ANALY­SIS
  4. FLUX BAL­ANCE ANALY­SIS
  5. EXPORT
  6. FIL­TERS
  7. TOPO­LOG­I­CAL INFOR­MA­TION GLYPHES
  8. Using Cytoscape with MetExplore


Top

Overview



Met­Ex­plore is a web-​server that allows to build, curate and analyse genome-​scale meta­bolic networks.

Met­Ex­plore stores meta­bolic net­works of 160 organ­isms into a rela­tional data­base (freely avail­able on demand). Infor­ma­tion about meta­bolic net­works mainly come from BioCyc-​like data­bases. Two BioCyc-​like data­bases con­tain infor­ma­tion about sev­eral organ­isms: Plant­Cyc and Meta­Cyc.
Met­Ex­plore con­tains also the infor­ma­tion about metabo­lites stored in Metabolome​.jp. Note that there is no infor­ma­tion about reac­tions in this data­base and is only use­ful to iden­tify com­pounds from masses.
Sev­eral genome-​scale mod­els designed for Flux Bal­ance Analy­sis have also been imported into Met­Ex­plore.
The table below gives details about the sources of the meta­bolic net­works present in MetExplore.


Var­i­ous fil­ters can be applied in Met­Ex­plore to restrict the scope of the study, for exam­ple by select­ing only par­tic­u­lar path­ways or by restrict­ing the net­work to the small-​molecule metab­o­lism. Met­Ex­plore also allows fil­ter­ing of cur­rency metabo­lites, cofac­tors or generic reac­tions, often sources of mis­in­ter­pre­ta­tions in meta­bolic graph analy­sis and which also bring sig­nif­i­cant quan­ti­ties of noise to the derived net­works. These fil­ters are described in the next sec­tion.



Met­Ex­plore is able to deal with data from metabolomics exper­i­ments by map­ping a list of masses or iden­ti­fiers onto fil­tered meta­bolic networks.


Met­Ex­plore pro­vides sev­eral func­tions based on the mod­el­ling of the meta­bolic net­work by a graph model.
Two func­tions are ded­i­cated to the iden­ti­fi­ca­tion of “weak” points in the meta­bolic net­work to help the design of new drugs. The first one iden­ti­fies the choke point reac­tions , defined as reac­tions that either uniquely con­sume a spe­cific sub­strate or uniquely pro­duce a spe­cific prod­uct. The sec­ond one iden­ti­fies the choke point metabo­lites , defined as metabo­lites that are either uniquely con­sumed by a spe­cific reac­tion or uniquely pro­duced by a spe­cific reac­tion.
The scope func­tion allows to com­pute the biosyn­thetic capac­ity of a set of metabo­lites. The reverse of the scope func­tion is also pos­si­ble in Met­Ex­plore by iden­ti­fy­ing the set of pre­cur­sors that lead to the pro­duc­tion of a given set of metabo­lites.
Met­Ex­plore allows to dis­play infor­ma­tion about the whole sets of metabo­lites present in a fil­tered meta­bolic net­work. Apart from clas­si­cal infor­ma­tion such as mass, for­mula, reac­tions and path­ways where each metabo­lite is involved, Met­Ex­plore dis­plays also the topo­log­i­cal prop­er­ties of the metabo­lite in the fil­tered meta­bolic net­work. It is then easy to know whether a metabo­lite is a source, an out­put or a choke point in the fil­tered meta­bolic network.

Based on the Sur­reyFBA library, Met­Ex­plore pro­poses sev­eral func­tions to per­form Flux Bal­ance Analy­sis (FBA). Com­pared to other flux analy­sis soft­wares, the most impor­tant con­tri­bu­tion of Met­Ex­plore is to make much eas­ier the tun­ing of flux para­me­ters.
The sim­plest FBA func­tion of Met­Ex­plore is to com­pute the opti­mal value of the objec­tive func­tion (a reac­tion or a set of reac­tions whose you want to min­i­mize or to max­i­mize the flux value).
You can per­form the same oper­a­toin but with adding some gene or reac­tions kos.
The flux vari­abil­ity analy­sis (FVA) returns for a set of selected reac­tion­s­the min­i­mum and the max­i­mum fluxes that allow the value of the objec­tive func­tion to be optimal.The result can be mapped on the com­plete meta­bolic net­work by launch­ing dire­clty Cytoscape from Met­Ex­plore.
Met­Ex­plore allows also to com­pare two FVA per­formed on two con­di­tions. Plot func­tions allow to exam­ine the effect of vary­ing one flux or two fluxes in the opti­mal value.
Essen­tial reac­tions allow to iden­tify key reac­tions in the real­i­sa­tion of the objec­tive func­tion.
At last, two FBA func­tions can be used in meta­bolic model cura­tion: the detec­tion of the live reac­tions, i.e. those able to carry a steady state flux con­sid­er­ing the flux con­straints and the iden­ti­fi­ca­tion of the orphan metabo­lites, i.e inter­nal metabo­lites used by less than two reactions.

While Met­Ex­plore makes avail­able draft meta­bolic recon­struc­tions, it allows also the build­ing and the cura­tion of meta­bolic net­works via a user inter­face that makes eas­ier sev­eral cura­tion tasks.
The Met­Ex­plore anno­ta­tion mod­ule allows the import, anno­ta­tion and mod­i­fi­ca­tion of meta­bolic net­works, thereby enabling to track miss­ing, false or sto­i­chio­met­ri­cally badly-​defined reac­tions, and ensur­ing highly con­sis­tent recon­structed mod­els. In addi­tion, Met­Ex­plore enables the cura­tion of meta­bolic mod­els by sev­eral users in a col­lab­o­ra­tive way. The meta­bolic net­works used to estab­lish mod­els can be dupli­cated from the Met­Ex­plore data­base or imported from exter­nal SBML files.


At last, the Met­Ex­plore user can export a fil­tered meta­bolic net­work into sev­eral graph mod­els (com­pound graph, reac­tion graph or bipar­tite graph). One can also upload his own SBML file and export it in graph mod­els.

Sources of the meta­bolic net­works stored in MetExplore

Data­base sourceVer­sionOrgan­isms
Genome-​scale mod­els imported from SBML Files AraGEM: Ara­bidop­sis thaliana (Gomes et al., 2009)
iRC1080: Chlamy­domonas rein­hardti (Chang et al., 2011)
iJO1366: Escherichia coli model (Orth et al., 2011)
Ec_​iAF1260 (flux1 and flux2): Escherichia coli mod­els (Feist et al., 2007)
iAC560: Leish­ma­nia major (Chavali et al.)
imm904: Sac­cha­romyces cere­visiae (Mo et al., 2009)
Bio­Cyc 17.5 (Dec, 29, 2013) Acine­to­bac­ter bau­man­nii 6013113
Acine­to­bac­ter bau­man­nii 6013150
Acine­to­bac­ter bau­man­nii 6014059
Acine­to­bac­ter bau­man­nii AB0057
Acine­to­bac­ter bau­man­nii ABAU557600
Acine­to­bac­ter bau­man­nii ACICU
Acine­to­bac­ter bau­man­nii ATCC 17978
Acine­to­bac­ter bau­man­nii ATCC 19606
Acine­to­bac­ter bau­man­nii AYE
Acine­to­bac­ter bau­man­nii SDF
Bos tau­rus
Escherichia coli K12
Homo sapi­ens
Meta­Cyc : con­tains infor­ma­tion about 1670 path­ways from more than 2100 dif­fer­ent organ­isms
Leish­ma­nia major Friedlin
Mus mus­cu­lus
Plas­mod­ium bergheistrain ANKA
Plas­mod­ium fal­ci­parum 3D7
Plas­mod­ium vivax Sal­vador I
Plas­mod­ium yoelii yoeliistr. 17XNL
Pseudovib­rio sp. FO-​BEG1
Sac­cha­romyces cere­visiae S288C
Try­panosoma brucei
Micro­Cyc Dec, 29, 2013 Agrobac­terium tume­fa­ciens C58
Bacil­lus amy­loliq­ue­fa­ciens FZB42
Bacil­lus anthracis Ames Ances­tor
Bacil­lus sub­tilis 168
Bacil­lus thuringien­sis serovar konkukian str. 9727
Bar­tonella bacil­li­formis KC583
Bar­tonella hense­lae Houston-​1
Bar­tonella quin­tana Toulouse
Bar­tonella tri­bo­co­rum CIP 105476
Blat­tabac­terium sp Blat­tella ger­man­ica Bge
Bradyrhi­zo­bium japon­icum USDA 110
Bradyrhi­zo­bium sp. BTAi1
Bru­cella meliten­sis bv 1 16M
Buch­n­era aphidi­cola APS (Acyrthosiphon pisum)
Buch­n­era aphidi­cola Bp (Baizon­gia pista­ciae)
Buch­n­era aphidi­cola Cc (Cinara cedri)
Buch­n­era aphidi­cola Sg (Schiza­phis graminum)
Burk­holde­ria mallei ATCC 23344
Campy­lobac­ter jejuni RM1221
Campy­lobac­ter jejuni subsp. jejuni str. NCTC 11168
Can­di­da­tus Bau­man­nia cicadellini­cola Hc
Can­di­da­tus Blochman­nia flori­danus
Can­di­da­tus Blochman­nia penn­syl­van­i­cus BPEN
Can­di­da­tus Car­sonella rud­dii PV
Can­di­da­tus Hamil­tonella defensa T5A (Acyrthosiphon pisum)
Can­di­da­tus Hodgkinia cicadi­cola Dsem
Can­di­da­tus Sul­cia muel­leri GWSS
Can­di­da­tus Sul­cia muel­leri SMD­SEM
Chlamy­dia tra­choma­tis D/​UW-​3/​CX
Corynebac­terium glu­tam­icum R
Cupri­avidus tai­wa­nen­sis LMG19424
Desul­fo­talea psy­chrophila Lsv54
Erwinia caro­tovora subsp. atrosep­tica SCRI1043
Escherichia alber­tii TW07627
Escherichia coli 042
Escherichia coli 1011
Escherichia coli 53638
Escherichia coli 536
Escherichia coli 55989
Escherichia coli APEC AGI-​5
Escherichia coli APEC O1
Escherichia coli ATCC 8739
Escherichia coli B171
Escherichia coli B7A
Escherichia coli BL21-Gold(DE3)pLysS AG
Escherichia coli B str. REL606
Escherichia coli CFT073
Escherichia coli DH1
Escherichia coli E110019
Escherichia coli E22
Escherichia coli E24377A
Escherichia coli ED1a
Escherichia coli ETEC H10407
Escherichia coli F11
Escherichia coli HS
Escherichia coli IAI1
Escherichia coli IAI39
Escherichia coli K-​12 BW2952
Escherichia coli K-​12 DH10B
Escherichia coli K12
Escherichia coli K-​12 sub­str. W3110
Escherichia coli LF82
Escherichia coli O103:H2 str. 12009
Escherichia coli O104:H4 LB226692
Escherichia coli O111:H– str. 11128
Escherichia coli O127:H6 E2348/​69
Escherichia coli O157:H7_EC4042
Escherichia coli O157:H7_EC4045
Escherichia coli O157:H7 EC4115
Escherichia coli O157:H7_EC4206
Escherichia coli O157:H7 EDL933
Escherichia coli O157:H7
Escherichia coli O157:H7_TW14588
Escherichia coli O26:H11 str. 11368
Escherichia coli S88
Escherichia coli SE11
Escherichia coli SE15
Escherichia coli SMS-​35
Escherichia coli SMS-​35
Escherichia coli UMN026
Escherichia coli UTI89
Escherichia fer­gu­sonii
Frankia alni ACN14a
Heli­cobac­ter pylori 26695
Heli­cobac­ter pylori Shi470
Kleb­siella pneu­mo­niae 342
Kleb­siella pneu­mo­niae NTUH-​K2044
Kleb­siella pneu­mo­niae pneu­mo­niae
Kleb­siella pneu­mo­niae subsp pneu­mo­niae MGH 78578
Lac­to­bacil­lus casei ATCC 334
Law­so­nia intra­cel­lu­laris PHE/​MN1-​00
Lis­te­ria mono­cy­to­genes EGD-​e
Mesorhi­zo­bium loti MAFF303099
Methy­lobac­terium extorquens AM1
Methy­lobac­terium extorquens DM4
Methy­lobac­terium extorquens PA1
Methy­lobac­terium pop­uli BJ001
Methy­lobac­terium radiotol­er­ans JCM 2831
Methy­lobac­terium sp. 446
Mycobac­terium smeg­ma­tis MC2 155
Mycobac­terium tuber­cu­lo­sis H37Rv
Mycoplasma gen­i­tal­ium G37
Mycoplasma hyop­neu­mo­niae 232
Mycoplasma hyop­neu­mo­niae 7448
Mycoplasma hyop­neu­mo­niae J
Neis­se­ria gon­or­rhoeae NCCP11945
Ori­en­tia tsut­sug­a­mushi Bory­ong
Pho­torhab­dus lumi­nescens TTO1
Pseudoal­teromonas halo­plank­tis TAC125
Pseudomonas aerug­i­nosa LESB58
Pseudomonas aerug­i­nosa PAO1
Pseudomonas aerug­i­nosa UCBPP-​PA14
Pseudomonas ento­mophila L48
Pseudomonas flu­o­rescens Pf0-​1
Pseudomonas flu­o­rescens Pf-​5
Pseudomonas flu­o­rescens SBW25
Pseudomonas putida F1
Pseudomonas putida GB-​1
Pseudomonas syringae pv. phase­oli­cola
Pseudomonas syringae pv. syringae
Ral­sto­nia eutropha H16
Ral­sto­nia eutropha H16
Ral­sto­nia eutropha JMP134
Ral­sto­nia pick­et­tii 12J
Ral­sto­nia solanacearum CFBP2957
Ral­sto­nia solanacearum CMR15
Ral­sto­nia solanacearum GMI10005
Ral­sto­nia solanacearum Ipo1609
Ral­sto­nia solanacearum K60
Ral­sto­nia solanacearum Molk2
Ral­sto­nia solanacearum Po82
Ral­sto­nia solanacearum PSI07
Ral­sto­nia solanacearum Y45
Ral­sto­nia syzy­gii R24
Rhi­zo­bium etli CFN 42
Rhi­zo­bium legu­mi­nosarum bv. viciae 3841
Rhodobac­ter sphaeroides 2.4.1
Rick­ettsia akari Hart­ford
Rick­ettsia bel­lii OSU 85389
Rick­ettsia mas­sil­iae MTU5
Rick­ettsia prowazekii Madrid E
Rick­ettsia rick­ettsii Iowa
Rick­ettsia rick­ettsii Sheila Smith
Rick­ettsia typhi Wilm­ing­ton
Sal­mo­nella enter­ica serovar Typhi
Shigella boy­dii Sb227
Shigella dysen­te­riae Sd197
Shigella flexneri 2a 2457T
Shigella flexneri 2a 301
Shigella flexneri 2a str. 301
Shigella flexneri 2a str. 301
Shigella flexneri 5 8401
Shigella son­nei Ss046
Sinorhi­zo­bium meliloti 1021
Sodalis glossini­d­ius mor­si­tans
Strep­to­coc­cus agalac­tiae 2603V/​R
Strep­to­coc­cus ther­mophilus LMD-​9
Thiomi­crospira cruno­gena XCL-​2
Vib­rio cholerae O1 bio­var eltor str. N16961
Wig­gleswor­thia glossini­dia endosym­biont of Glossina bre­vipalpis
Wol­bachia pip­i­en­tis wBm pip­i­en­tis
Wol­bachia pip­i­en­tis wMel
Wolinella suc­cino­genes DSM 1740
Xylella fas­tidiosa 9a5c
Yersinia pestis CO92
Yersinia pseudo­tu­ber­cu­lo­sis YPIII
Metabolome​.jp May 2006 Data­base for metabo­lites. Does not con­tain infor­ma­tion about reactions.
Plant Meta­bolic Network Dec, 29, 2013 Ara­bidop­sis thaliana col
Brachy­podium dis­tachyonNew !
Bras­sica rapaNew !
Car­ica papaya
Chlamy­domonas rein­hardtii
Glycine max
Hordeum vul­gareNew !
Mani­hot escu­lenta escu­lenta
Oryza sativaNew !
Pan­icum vir­ga­tumNew !
Physcomitrella patens
Plant­Cyc (Meta­bolic infor­ma­tion about over 500 path­ways in more than 250 plant species)
Pop­u­lus tri­chocarpa
Selaginella moel­len­dorf­fii
Setaria ital­icaNew !
Sorghum bicolorNew !
Vitis vinifera
Zea mays mays

Sources of the meta­bolic net­works stored in MetExplore


Top

METABOLOME MAP­PING

Tuto­r­ial: Metabolome map­ping with MetExplore

In Met­Ex­plore, three ways exist to map metabolome onto fil­tered meta­bolic net­works, either by iden­ti­fy­ing metabo­lites from a list of masses or by using a list of metabo­lite iden­ti­fiers or by using a list of metabo­lite names.

Top

Map Masses

To launch this func­tion, click “Map Masses” in the top menu of MetExplore.



Once an organ­ism is selected, the fol­low­ing page is dis­played in the main frame of Met­Ex­plore. Here, the inline exam­ple is loaded.



The Met­Ex­plore user can upload his (her) data by two ways: by upload­ing a file or by copy­ing and past­ing data. In the two cases, the for­mat is the same.

The first col­umn always spec­i­fies the masses and the fol­low­ing columns cor­re­spond to numer­i­cal val­ues (e.g., peak inten­si­ties for metabo­lites iden­ti­fied under dif­fer­ent con­di­tions or reten­tion times).

Each col­umn must be sep­a­rated by a tab­u­la­tion and each line must have the same num­ber of columns. A line begin­ning by the char­ac­ter “#” is con­sid­ered as a com­ment and is not taken into account in the map­ping.

By default, the first line does not cor­re­spond to head­ers of the columns. If the first col­umn cor­re­sponds to head­ers, answer “Yes” to the ques­tion “Is the first line cor­re­spond to the head­ers of the columns?”. The col­umn names will be dis­played in the results.
Note: At the moment, Met­Ex­plore does not allow a num­ber of columns greater than 20.

Click on “Load the exam­ple” to have an exam­ple of the required for­mat.

Each mass of the input data will be com­pared to the masses of the metabo­lites present in the fil­tered meta­bolic net­work of the selected organ­ism. Depend­ing on the pre­ci­sion of the input masses, the Met­Ex­plore user can allow an error indi­cated in ppm (parts-​per-​million).

The fil­ters described below can be dis­played by click­ing on the top tabs. Be care­ful, some fil­ters are acti­vated by default.

The “sub­mit” but­ton launches the map­ping. Depend­ing on the num­ber of input masses and the size of the meta­bolic net­work, it may take some time but should not exceed five min­utes.

Results are dis­played in the main frame of Met­Ex­plore as shown in the image below.

Met­Ex­plore Mass Map­ping main page result



The table dis­played cor­re­sponds to the masses iden­ti­fied in the fil­tered meta­bolic net­work of the selected organ­ism. Each col­umn is sortable by click­ing on the arrows in the headers.

  • The first col­umn con­tains the names of the com­pounds iden­ti­fied. Each name is linked to the cor­re­spond­ing page in the source database.
  • The sec­ond col­umn con­tains the iden­ti­fiers of the com­pounds iden­ti­fied in the source database.
  • The third col­umn con­tains the masses identified.
  • The fourth col­umn con­tains the for­mula of the com­pounds identified.
  • The fifth col­umn con­tains the path­ways where the metabo­lite is involved in the fil­tered meta­bolic net­work of the selected organ­ism. Each path­way is linked to the cor­re­spond­ing page in the source database.
  • The sixth col­umn con­tains the glyphes describ­ing the topo­log­i­cal infor­ma­tion of the metabo­lites in the fil­tered meta­bolic network.
  • The fol­low­ing columns cor­re­sponds to the numer­i­cal val­ues given as input by the user. To make eas­ier the iden­ti­fi­ca­tion of low or high val­ues, each value is col­ored depend­ing on the quar­tile of the whole set of val­ues of the same input col­umn it belongs. The col­ors are shown in the table below.

The table ca be sorted using the val­ues in columns by click­ing on or on .
In order to make the table clearer, columns can be removed by click­ing on .
The names of the columns removed appear above the table. By click­ing on a col­umn name, the cor­re­spond­ing col­umn is dis­played again.

First quar­tile Sec­ond quartile Third quar­tile Fourth quar­tile

Back­ground color of the numer­i­cal values

On the top of the frame, there are five tabs. The first one cor­re­sponds to the table of results that we just described. The sec­ond tab cor­re­sponds to the table of masses that have not been iden­ti­fied in the fil­tered meta­bolic network.

Met­Ex­plore Mass Map­ping: Not iden­ti­fied masses



The third tab resumes the fil­ters applied on the meta­bolic network.

Met­Ex­plore Mass Map­ping: Resume of the filters



The fourth tab dis­plays the links to the SBML, graph and attribute files cor­re­spond­ing to the fil­tered meta­bolic net­work and the results of the analysis.

Met­Ex­plore Mass Map­ping: net­work and attribute files



Net­work files
SBML is an xml for­mat ded­i­cated to the meta­bolic net­works. Some addi­tional infor­ma­tions are stored in the SBML files gen­er­ated by Met­Ex­plore. Visit the SBML sec­tion to have a descrip­tion of the Met­Ex­plore SBML for­mat. SBML files can be directly visu­alised in Cytoscape and used by numer­ous soft­wares to per­form fur­ther meta­bolic analy­ses.
The fil­tered meta­bolic net­work is also down­load­able as a bipar­tite graph in Cytoscape for­mat. This for­mat con­sists in a list of edges. Read the graph sec­tion to have a def­i­n­i­tion of the bipar­tite graph and a descrip­tion of the for­mat.

Attribute files
Two attribute files are gen­er­ated by the map­ping: one con­tains gen­eral attrib­utes of the fil­tered meta­bolic net­work and the other one con­tains infor­ma­tions about metabo­lites iden­ti­fied from the input masses. Both are in Cytoscape for­mat and can be loaded in the soft­ware. Please read the Cytoscape sec­tion to have a descrip­tion of the gen­eral attrib­utes table. Note that all the attrib­utes are auto­mat­i­cally loaded when Cytoscape is lauc­n­hed from Met­Ex­plore (see next sec­tion).

Each line of the detected com­pounds table cor­re­sponds to a metabo­lite iden­ti­fied by the map­ping. The columns cor­re­spond to the fol­low­ing attributes:

  • ID: the iden­ti­fier of the metabo­lite in SBML for­mat,
  • iden­ti­fied: always equals to iden­ti­fied, allows to eas­ily locate the iden­ti­fied metabo­lites in Cytoscape and is used by the Gap­Filler plu­gin (see Cytoscape section),
  • sbml name: the com­mon name of the reac­tion or of the metabolite,
  • the fol­low­ing columns cor­re­spond to the numer­i­cal val­ues given by the user.

Note: if the same metabo­lite cor­re­sponds to sev­eral input masses, the numer­i­cal val­ues writ­ten in the attribute file are those cor­re­spond­ing to the last mass in the input file.



Visu­al­i­sa­tion of the map­ping results in Cytoscape
The fifth tab “Launch Cytoscape allows to visu­alise the map­ping results into Cytoscape.

Met­Ex­plore Mass Map­ping: Launch Cytoscape to visu­alise the results



Please read the “Using Cytoscape with Met­Ex­plore” sec­tion to know the required speci­fici­ties and the con­ven­tions used to rep­re­sent a Met­Ex­plore meta­bolic net­work into Cytoscape. Once Cytoscape launched, the meta­bolic net­work and all the cor­re­spond­ing attrib­utes are loaded. The image below rep­re­sents a Met­Ex­plore net­work dis­played in Cytoscape (with the organic lay­out) with the iden­ti­fied metabo­lites highlighted.

Met­Ex­plore Mass Map­ping: Visu­al­i­sa­tion of the results in Cytoscape



The metabo­lites iden­ti­fied by the mass map­ping are larger than the oth­ers and are col­ored in dark blue. All the attrib­utes we described above are directly avail­able in Cytoscape.

Met­Ex­plore Mass Map­ping: attrib­utes of the net­work and of the map­ping into Cytoscape



Sub-​network extrac­tion from iden­ti­fied metabo­lites As the rep­re­sen­ta­tion of the iden­ti­fied metabo­lites in the whole net­work remains dif­fi­cult, we devel­oped a Cytoscape plu­gin to com­pute the sub-​networks link­ing them. For each iden­ti­fied metabo­lite we look at all of the reac­tions that use this given metabo­lite as a sub­strate. If all of the other sub­strates of this reac­tion are present in the dataset, we con­sider all of the prod­ucts of the reac­tion as poten­tial can­di­dates. For each of these can­di­dates we search the list of all reac­tions for which the can­di­date is a sub­strate. If for at least one of these reac­tions all required sub­strates and prod­ucts are present in the dataset, we con­sider that the metabo­lite in ques­tion is a rel­e­vant can­di­date. Finally, we add both the can­di­date metabo­lite the asso­ci­ated reac­tions which helped us to localise it in the sub-​network of the pre­vi­ous iter­a­tion.
To launch the plu­gin, go in the plu­g­ins menu and click on “Gap Filler”.

Met­Ex­plore Mass Map­ping: launch Gap Filler



The sub-​networ ks are directly com­puted and dis­played in Cytoscape.

Met­Ex­plore Mass Map­ping: Gap Filler sub-​network extraction



Top


Map Iden­ti­fiers

To launch this func­tion, click on “Map Iden­ti­fiers” in the top menu of MetExplore.

Launch Met­Ex­plore Iden­ti­fier Mapping



This func­tion is sim­i­lar to the mass map­ping func­tion of Met­Ex­plore. The input is not masses but directly iden­ti­fiers of metabo­lites. The Met­Ex­plore Iden­ti­fier Map­ping page looks like the Met­Ex­plore Mass Map­ping. The for­mat used to upload the data is exactly the same, except for the first col­umn which cor­re­sponds to metabo­lite iden­ti­fiers. Please refer to the Mass Map­ping sec­tion to have a descrip­tion of the inputs.

Met­Ex­plore Iden­ti­fier Mapping



The result page is also exactly the same as the Mass Map­ping result page, except the “Not Iden­ti­fied” tab that does not exist in the Iden­ti­fier Map­ping result page. The iden­ti­fiers not found in the fil­tered meta­bolic net­work are just indi­cated before dis­play­ing the results. Please refer to the Mass Map­ping sec­tion to have a descrip­tion of the results.




Top

Map Names

To launch this func­tion, click on “Map Names” in the top menu of MetExplore.

Launch Met­Ex­plore Name Mapping



This func­tion is also sim­i­lar to the mass map­ping func­tion of Met­Ex­plore. The input is not masses but directly names of metabo­lites. The Met­Ex­plore Names Map­ping page looks like the Met­Ex­plore Mass Map­ping. The for­mat used to upload the data is exactly the same, except for the first col­umn which cor­re­sponds to metabo­lite names. The input names can include reg­u­lar expres­sions such as described here. The reg­u­lar expres­sion to indi­cate the beg­gin­ing (^) and the end of the pat­tern ($) are directly included dur­ing the map­ping. Please refer to the Mass Map­ping sec­tion to have a descrip­tion of the inputs.

Met­Ex­plore Names Mapping



The result page is also exactly the same than the Mass Map­ping result page, except the “Not Iden­ti­fied” tab that does not exist in the Names Map­ping result page. The names not found in the fil­tered meta­bolic net­work are just indi­cated before dis­play­ing the results. The ambigu­ous names, i.e. that cor­re­spond to sev­eral metabo­lites are also indi­cated before the map­ping. Please refer to the Mass Map­ping sec­tion to have a descrip­tion of the results.


META­BOLIC GRAPH ANALYSIS

Apart from the metabolome map­ping, Met­Ex­plore includes sev­eral graph analy­ses that can be applied in the meta­bolic net­works stored into Met­Ex­plore. These func­tions are:

Top

Choke Point Reactions

Tuto­r­ial: iden­ti­fi­ca­tion of choke point reac­tions with MetExplore

This func­tion allows the user to iden­tify choke point reac­tions in a fil­tered meta­bolic net­work. A choke point reac­tion is defined as a reac­tion that either uniquely con­sumes a spe­cific sub­strate or uniquely pro­duces a spe­cific prod­uct. Iden­ti­fy­ing such reac­tions can be used for instance to deter­mine drug tar­gets.

To launch this func­tion, click on “Choke Point Reac­tions” in the top menu of MetExplore.

Launch Met­Ex­plore Choke Point Reac­tions Identification



Select a dataset, spec­ify fil­ters if you want to change those by default (read the Fil­ters sec­tion to have a descrip­tion of the fil­ters avail­able in Met­Ex­plore), and click on “Com­pute Choke Points”.

Met­Ex­plore Choke Point Reac­tions Identification



After a few sec­onds, a result table is displayed.

Met­Ex­plore Choke Point Reac­tions Iden­ti­fi­ca­tion: table of results



  • The first col­umn cor­re­sponds to the reac­tion name and is linked to the cor­re­spond­ing page in the source database.
  • The sec­ond col­umn cor­re­sponds to the EC num­ber of the reaction.
  • The third col­umn cor­re­sponds to the for­mula of the reaction.
  • The fourth col­umn cor­re­sponds to a flag indi­cat­ing if the reac­tion is present or not in the human meta­bolic net­work. If the BioSource selected come from a BioCyc-​like data­base, the reac­tion iden­ti­fier is searched into the com­plete meta­bolic net­work from Human­Cyc. This flag is inter­est­ing in the con­text of the design of drugs against a par­a­site hosted by humans. In this case, it is impor­tant to tar­get reac­tions not present in the human.
  • The fourth col­umn cor­re­sponds to the path­ways in which the reac­tion is involved in the fil­tered meta­bolic net­work. Each path­way is linked to the cor­re­spond­ing page in the source database.

Below the table, the num­ber of choke points found in the fil­tered meta­bolic net­work, the links to attribute tables and the link to launch Cytoscape with the results attrib­utes are displayed.

Met­Ex­plore Choke Point Reac­tions Iden­ti­fi­ca­tion: links to attribute files and to Cytoscape



Please read the SBML For­mat Sec­tion to have a descrip­tion of the SBML files gen­er­ated by Met­Ex­plore.
The gen­eral attrib­utes table con­tained in the file “Attrib­utes table” are described here.

The file “Choke Reac­tions in Cytoscape for­mat” con­tains the list of all the reac­tions that are iden­ti­fied as choke point reac­tions. This file is in the Cytoscape for­mat designed to store the attrib­utes. The first line con­tains the name of the attribute (here “choke”). The fol­low­ing lines con­tains the iden­ti­fier of a choke point reac­tion and the value “choke” sep­a­rated by ” = “.
The file “Pres­ence of the choke reac­tions in human in Cytoscape for­mat” con­tains, for all the iden­ti­fied choke points, a flag indi­cat­ing if the reac­tion is present in the human meta­bolic net­work. The first line con­tains the name of the attribute (here “Choke­PointsIn­Hu­man”) and the fol­low­ing lines con­tain the iden­ti­fier of a choke point reac­tion and the value F (false) or T (true) to indi­cate its pres­ence in human sep­a­rated by ” = “.

When the Met­Ex­plore user launches Cytoscape from this page, the fil­tered meta­bolic net­work and all the attrib­utes described above are directly loaded. It is thus pos­si­ble to adapt the visu­al­i­sa­tion style of Cytoscape to high­light the infor­ma­tion about choke point reac­tions. Please refer to the Cytoscape doc­u­men­ta­tion about visual styles to know how to per­form visu­al­i­sa­tion tunings.



Top

Choke Point Metabolites

Tuto­r­ial: Choke Point Metabo­lites Identification

The choke point metabo­lites are defined as metabo­lites either uniquely con­sumed by a spe­cific reac­tion or uniquely pro­duced by a spe­cific reac­tion.

To launch this func­tion, click on “Choke Point Metabo­lites” on the top menu of MetExplore.

Launch Met­Ex­plore Choke Point Metabo­lites Identification



Once an organ­ism selected and the meta­bolic net­work fil­tered (read the Fil­ters sec­tion), click on “Com­pute Choke Point Metabolites”.

Choke Point Metabo­lites Identification



After a few sec­onds, a result table is displayed.

Met­Ex­plore Choke Point Metabo­lites Result Page



The columns of the table cor­re­sponds to:

  • the name of the metabo­lite linked to the cor­re­spond­ing page in the source database,
  • its chem­i­cal formula,
  • a flag indi­cat­ing its pres­ence in human (we cur­rently con­sider the com­plete meta­bolic net­work from HumanCyc),
  • a glyph describ­ing the topo­log­i­cal infor­ma­tion of the metabo­lite in the fil­tered meta­bolic net­work (Glyphes descrip­tion),
  • the list of path­ways where the metabo­lite is involved.

After the result table, the total num­ber of choke point metabo­lites is indi­cated, links to the SBML and attrib­utes files and to the Cytoscape launcher are displayed.

Met­Ex­plore Choke Point Metabo­lites Result Page



Please read the SBML For­mat Sec­tion to have a descrip­tion of the SBML files gen­er­ated by Met­Ex­plore. The gen­eral attrib­utes table con­tained in the file “Attrib­utes table” are described here.
The file “Choke point metabo­lites in Cytoscape for­mat” con­tains the list of all the metabo­lites that are iden­ti­fied as choke point com­pounds. This file is in the Cytoscape for­mat designed to store the attrib­utes. The first line con­tains the name of the attribute (here “chokeCpd”). The fol­low­ing lines con­tains the iden­ti­fier of a choke point metabo­lite and the value “choke” sep­a­rated by ” = “.
The file “Pres­ence of the choke point metabo­lites in human in Cytoscape for­mat” con­tains, for all the iden­ti­fied choke points, a flag indi­cat­ing if the metabo­lite is present in the human meta­bolic net­work. The first line con­tains the name of the attribute (here “Choke­PointsIn­Hu­man”) and the fol­low­ing lines con­tain the iden­ti­fier of a choke point metabo­lite and the value F (false) or T (true) to indi­cate its pres­ence in human sep­a­rated by ” = “.

When the Met­Ex­plore user launches Cytoscape from this page, the fil­tered meta­bolic net­work and all the attrib­utes described above are directly loaded. It is thus pos­si­ble to adapt the visu­al­i­sa­tion style of Cytoscape to high­light the infor­ma­tion about choke point metabo­lites. Please refer to the Cytoscape doc­u­men­ta­tion about visual styles to know how to per­form visu­al­i­sa­tion tunings.



Top

Scope of Metabolites

Tuto­r­ial: com­put­ing the scope of a set of metabo­lites with MetExplore

The scope of metabo­lites was first defined by Han­dorf et al. in this paper and allows to iden­tify the poten­tial biosyn­thetic capac­ity of a set of metabo­lites. The scope of a set of metabo­lites (so-​called seeds) is defined as the sum of all metabo­lites that the seeds are able to pro­duce using the reac­tions avail­able in an organ­ism. On the con­trary of short­est paths com­puted in sim­ple graphs, the scope con­cept takes into account the avail­abil­ity of all the sub­strates to use a reac­tion. The scope is com­puted in an iter­a­tive way, called expan­sion process. At each step, the reac­tions using inputs are checked: if all the sub­strates are in the set of seeds, then they are fired and all their prod­ucts join the set of seeds. The process stops when no addi­tional reac­tion can be fired. The metabo­lites con­tained in the final set of seeds rep­re­sent the scope of the ini­tial set of seeds.

To launch this func­tion, click on “Scope of Metabo­lites” in the left Menu of MetExplore.

Launch Met­Ex­plore Scope Function



After select­ing a BioSource, a table is dis­played and the user can select the seeds and the boot­strap metabolites.These are not used as seeds but allow to fire a reac­tion dur­ing the expan­sion process if all its sub­strates are seeds or boot­strap metabo­lites. These metabo­lites often cor­re­spond to cofac­tor metabo­lites as ATP or NADH, con­sid­ered as avail­able in the meta­bolic net­work but that the user doesn’t want to con­sider as seeds.

Met­Ex­plore Scope Function



Please read this sec­tion to have a descrip­tion of the fil­ters that can be applied on the meta­bolic net­work of the selected organ­ism.

Click on “Sub­mit” to com­pute the scope of the metabo­lites given as seeds in the input file. After a few sec­onds, a result page is displayed.

Met­Ex­plore Scope Result Page



The columns of the result page cor­re­spond to:

  • the name of a metabo­lite pro­duced dur­ing the expan­sion process linked to the data­base source,
  • its chem­i­cal formula,
  • the dis­tance of the metabo­lite from the input com­pounds (seeds), i.e. the num­ber of iter­a­tions per­formed before vis­it­ing the metabo­lite dur­ing the expan­sion process,
  • a glyphe describ­ing the topo­log­i­cal infor­ma­tion of the metabo­lite in the fil­tered meta­bolic net­work (read this sec­tion to have the mean­ing of each glyphe).

After the table, the links to the SBML files, attrib­utes files and to the Cytoscape launcher are displayed.

Met­Ex­plore Scope Result Page



Please read the SBML For­mat Sec­tion to have a descrip­tion of the SBML files gen­er­ated by Met­Ex­plore. The gen­eral attrib­utes table con­tained in the file “Attrib­utes table” are described here.

Apart from the fil­tered meta­bolic net­work of the selected organ­ism, the meta­bolic net­work cor­re­spond­ing to the reac­tions used and the metabo­lites pro­duced dur­ing the expan­sion process is avail­able in SBML for­mat by click­ing on the link “SBML file cor­re­spond­ing to the scope net­work”.
The file “Scope attrib­utes in Cytoscape for­mat” con­tains the num­ber of iter­a­tions used to fire metabo­lites or reac­tions dur­ing the expan­sion process. The first line con­tains the name of the attribute (step). Each fol­low­ing line con­tains the iden­ti­fier of a metabo­lite or of a reac­tion and the num­ber of iter­a­tions sep­a­rated by ” = “.

When the Met­Ex­plore user launches Cytoscape from this page, the fil­tered meta­bolic net­work, the scope net­work and all the attrib­utes described above are directly loaded. It is thus pos­si­ble to adapt the visu­al­i­sa­tion style of Cytoscape to high­light the infor­ma­tion about scope. Please refer to the Cytoscape doc­u­men­ta­tion about visual styles to know how to per­form visu­al­i­sa­tion tunings.



Top

Pre­cur­sors of Metabolites

Iden­ti­fi­ca­tion of the pre­cur­sors of a set of metabo­lites with MetExplore

This func­tion allows to com­pute the set of metabo­lites, called pre­cur­sors, suf­fi­cient to pro­duce a set of tar­get metabo­lites. This is com­puted by the inverse of the expan­sion process described in the Scope func­tion. The process starts with a set of given tar­get metabo­lites. At each step, the reac­tions pro­duc­ing tar­gets are checked, then they are fired and all their sub­strates join the set of tar­gets. The process stops when no addi­tional reac­tion can be fired. A metabo­lite is defined as a poten­tial pre­cur­sor of the set of tar­get metabo­lites if it is not pro­duced by any reac­tion or pro­duced only by one reversible reac­tion, and if there is a path between this com­pound and any of the tar­get com­pounds selected.
To launch this func­tion, click on “Pre­cur­sors of Metabo­lites” in the top Menu of Met­Ex­plore.

Once selected a BioSource, the user can select the tar­gets among the metabo­lites present in the BioSource.
An option allows to dis­play only the set of pre­cur­sors or all the reac­tions and metabo­lites used dur­ing the process. The image below dis­plays the main page of the Pre­cur­sors func­tion with the inline exam­ple loaded.

Met­Ex­plore Pre­cur­sors Page



Once an organ­ism selected and the meta­bolic net­work fil­tered (read the Fil­ters sec­tion) and the tar­get com­pounds given, click on “sub­mit”. After a few sec­onds, a result page is displayed.

Met­Ex­plore Pre­cur­sors Result Page



The columns of the result table cor­re­spond to:

  • the name of metabo­lites vis­ited dur­ing the process, the pre­cur­sors are marked with a red star,
  • the chem­i­cal for­mula of the metabolites,
  • the dis­tance of the metabo­lite to the clos­est tar­get metabo­lite, i.e. the num­ber of iter­a­tions per­formed before vis­it­ing the metabo­lite dur­ing the process,
  • a glyphe describ­ing the topo­log­i­cal infor­ma­tion of the metabo­lite in the fil­tered meta­bolic net­work (read this sec­tion to have the mean­ing of each glyphe).

After the table, the links to the SBML files, attrib­utes files and to the Cytoscape launcher are displayed.

Met­Ex­plore Pre­cur­sors Result Page



Please read the SBML For­mat Sec­tion to have a descrip­tion of the SBML files gen­er­ated by Met­Ex­plore. The gen­eral attrib­utes table con­tained in the file “Attrib­utes table” are described here.

Apart from the fil­tered meta­bolic net­work of the selected organ­ism, the meta­bolic net­work cor­re­spond­ing to the reac­tions used and the metabo­lites pro­duced dur­ing the expan­sion process is avail­able in SBML for­mat by click­ing on the link “SBML file cor­re­spond­ing to the back­track net­work”.
The file “Back­track attrib­utes in Cytoscape for­mat” con­tains the num­ber of iter­a­tions used to fire metabo­lites or reac­tions dur­ing the back­track­ing process. The first line con­tains the name of the attribute (step). Each fol­low­ing line con­tains the iden­ti­fier of a metabo­lite or of a reac­tion and the num­ber of iter­a­tions sep­a­rated by ” = “.

When the Met­Ex­plore user launches Cytoscape from this page, the fil­tered meta­bolic net­work, the back­track net­work and all the attrib­utes described above are directly loaded. It is thus pos­si­ble to adapt the visu­al­i­sa­tion style of Cytoscape to high­light the infor­ma­tion about pre­cur­sors. Please refer to the Cytoscape doc­u­men­ta­tion about visual styles to know how to per­form visu­al­i­sa­tion tunings.



Top

Metabolome prop­er­ties

This Met­Ex­plore func­tion allows to get infor­ma­tions about all the metabo­lites present in the fil­tered meta­bolic net­work of a selected organ­ism.
To launch the func­tion, click on “Metabolome graph prop­er­ties” in the left menu of Met­Ex­plore. Once an organ­ism selected and the meta­bolic net­work fil­tered (read the Fil­ters sec­tion), click on “Get the list of metabo­lites in the fil­tered meta­bolic net­work”. After a few sec­onds, a result page is displayed.

Met­Ex­plore Metabolome properties



The columns of the result table cor­re­spond to:

  • the name of the metabo­lite linked to the source database,
  • its iden­ti­fier in the source database,
  • its chem­i­cal formula,
  • its mol­e­c­u­lar weight,
  • a glyphe describ­ing the topo­log­i­cal infor­ma­tion of the metabo­lite in the fil­tered meta­bolic net­work (read this sec­tion to have the mean­ing of each glyphe)
  • the reac­tions using it as substrate,
  • the reac­tions using it as product,
  • the path­ways where it is involved.



Flux Bal­ance Analysis

Top

FBA returns a flux dis­tri­b­u­tion that opti­mizes a meta­bolic func­tion con­sid­er­ing con­straints on some fluxes. Please read this paper to have an overview of what is FBA.

Data require­ments

In Met­Ex­plore, Flux Bal­ance Analy­ses can be applied either on meta­bolic net­works stored in the Met­Ex­plore data­base or from SBML files. In the two cases, it’s impor­tant to well under­stand what is com­puted dur­ing FBA and so what must be present in the data.
The main assump­tion of FBA is that the sys­tem is in steady state. This means that inside the sys­tem, every quan­tity of metabo­lite that is pro­duced must be con­sumed. To mimic the exchanges with the exter­nal medium, the model has to con­tain exchange reac­tions that make metabo­lites to get in or out of the model. There are two ways to define the exchange reac­tions. The first one is to declare exter­nal metabo­lites, i.e. metabo­lites that are out­side the model and don’t have to be bal­anced. For instance, in the fig­ure below, R1 and R3 are exchange reac­tions that allow A to go in the model and B to go out of the model. This allows to tune the flux con­straints of the import of A and of the secre­tion of B.
In the SBML file, the exter­nal metabo­lites are indi­cated by the bound­aryCon­di­tion attribute that equals to true.

Bound­ing a meta­bolic model with exter­nal metabolites



In this exam­ple, the SBML file would be (see SBML spec­i­fi­ca­tions):

SBML file of a model bounded with exter­nal metabolites



The bound­ary con­di­tions of the metabo­lites can be changed in the BioSources for which the Met­Ex­plorer user has the read/​write rights (see the MetAn­not sec­tion).
Note that spec­i­fy­ing exter­nal metabo­lites can be used for metabo­lites for which the model is not able to bal­ance.

In a sec­ond way, the exchange reac­tions can be defined with an empty side (left or right) as in the fig­ure below. This has the advan­tage to avoid the dec­la­ra­tion of exter­nal metabolites.

Bound­ing a meta­bolic model with­out exter­nal metabolites



In this exam­ple, the SBML file would be:

SBML file of a model bounded with­out exter­nal metabolites



In the FBA met­ex­plore inter­face, it is pos­si­ble to spec­ify the lower and upper bounds of each reac­tion flux. How­ever, this infor­ma­tion can be already present in the SBML file. For instance, in our small meta­bolic net­work, if we want to limit the input flux of A from 10 mmol/​gDW/​hr to 15 mmol/​gDW/​hr, we add in the reac­tion R1 the fol­low­ing lines:

Flux bounds of a reac­tion in a SBML file

The unit (here mmol_​per_​gDW_​per_​hr) must be spec­i­fied in the begin­ning of the SBML file:

Unit def­i­n­i­tion in a SBML file

The flux bounds of the reac­tions can also be changed in the BioSources for which the Met­Ex­plorer user has the read/​write rights (see the MetAn­not sec­tion).
Be care­ful, in most of the SBML meta­bolic mod­els, exchange reac­tion are always writ­ten in the direc­tion in → out. This means that the input fluxes must be neg­a­tive so that the metabo­lites can go inside the model.
To be read­able by Met­Ex­plore, the gene asso­ci­a­tions and the path­ways must be writ­ten with one of these two for­mats in the SBML reac­tion element:

Gene asso­ci­a­tion and path­ways in SBML. First way.

Gene asso­ci­a­tion and path­ways in SBML. Sec­ond way.

Tun­ing FBA parameters

To Launch Flux Bal­ance Analy­ses, click on “Flux Bal­ance Analy­sis” on the top menu of MetExplore.

Launch FBA in MetExplore

The Met­Ex­plore user has the choice to load a meta­bolic net­work from the Met­Ex­plore data­base or from a SBML file.

Be care­ful, in the first case, the meta­bolic net­works stored in Met­Ex­plore are not directly usable for FBA (see Data require­ments above).
In the sec­ond case, please valid your sbml file using the SBML val­ida­tor before start­ing an analy­sis.

What­ever the source of the meta­bolic net­work selected, a table is dis­played when you click on “Fill FBA parameters”.

FBA tun­ing table in MetExplore

This table will allow to tune the flux con­straints, the gene asso­ci­a­tions, to rede­fine reac­tion for­mu­las, to spec­ify the objec­tive func­tion, and to select reac­tions used in flux analy­ses.
Each line cor­re­sponds to a reac­tion of the selected meta­bolic net­work.
The columns cor­re­spond to (the editable columns are high­lighted in green):

  1. Reac­tion names
  2. Reac­tion identifiers
  3. Flux lower bounds
  4. Flux upper bounds
  5. Par­tic­i­pa­tion to the objec­tive function.
    The objec­tive func­tion is a reac­tion of a set of reac­tions for which you want to max­i­mize the sum of the fluxes. An objec­tive func­tion can be writ­ten as a lin­ear com­bi­na­tion of fluxes such as 2 R1 + 3 R3. In this col­umn, the Met­Ex­plore user spec­i­fies the coef­fi­cients of each reac­tion involved in the objec­tive func­tion. To min­i­mize a flux, indi­cate a neg­a­tive coefficient.
  6. Type of reac­tions:
    • Input Exchange reac­tions (key­word: Exchange_​in). The exchange reversible reac­tions are also in this category.
    • Out­put Exchange reac­tions (key­word: Exchange_​out)
    • Inac­tive Exchange reac­tions (key­word: Exchange_​ko): exchange reac­tions for which the lower and upper bounds equal to 0
    • Trans­port reac­tions (key­word: Transport)
    • Inac­tive trans­port reac­tions (key­word: Transport_​ko): trans­port reac­tions for which the lower and upper bounds equal to 0
    • Inter­nal reac­tions (key­word: Inter­nal): reac­tions involv­ing only inter­nal (bal­anced) metabolites
    • Inac­tive Inter­nal reac­tions (key­word: Internal_​ko): inter­nal reac­tions for which the lower and upper bounds equal to 0
  7. Reac­tion for­mula with metabo­lite names. This allows to query the names of the metabo­lites (often more mean­ing­ful than the iden­ti­fiers) to find a reaction
  8. Reac­tion for­mula with metabo­lite identifiers.
    Since this col­umn is editable, each reac­tion for­mula can be changed before launch­ing analyses
  9. Gene asso­ci­a­tion.
    This col­umn con­tains the asso­ci­a­tions of genes needed to acti­vate the reac­tion. We use the way to write gene asso­ci­a­tions described in the Cobra Tool­Box. “Boolean rules for each reac­tion describ­ing the gene-​reaction rela­tion­ship. For exam­ple ‘gene1 and gene2’ indi­cate that the two gene prod­ucts are part of a enzyme com­ples whereas ‘gene1 or gene2’ indi­cate that the two gene prod­ucts are isozymes that cat­alyze the same reaction.”
  10. Reversibil­ity: T for true, F for false
  11. Path­ways
  12. Reac­tion selection.
    Check the boox to select reac­tions that can be used for some FBA prob­lems (see below)

Each col­umn is sortable and the seach box above the table can be used to fil­ter the table with key words. For instance, you can search for the reac­tions that con­tain the metabo­lite atp, that are involved in gly­col­y­sis, or that are only exchange reac­tions (in this case, use the key words men­tioned above).

By default, only 10 reac­tions are dis­played. With the pager below the table, the Met­Ex­plore user can browse the pages and select the num­ber of dis­played reactions.

Pager of the FBA tun­ing table in MetExplore

Below the tun­ing table, the Met­Ex­plore user selects the prob­lem type and fills the required parameters.

Selec­tion of the FBA prob­lem type in MetExplore



Com­pute the opti­mal objec­tive func­tion value

Tuto­r­ial: how to com­pute the opti­mal objec­tive func­tion value?

After spec­i­fy­ing the objec­tive func­tion, click on the check box cor­re­spond­ing to the “Opti­mal objec­tive func­tion value” and click on Sub­mit.
Met­Ex­plore dis­plays the opti­mal objec­tive func­tion value com­puted by the sur­reyFBA library.

Result of com­put­ing the opti­mal objec­tive func­tion value

Knock Out Flux Bal­ance Analysis

Tuto­r­ial : How to per­form knock-​out analy­sis with MetExplore?

This func­tion is the same than the pre­vi­ous one, excepted that the Met­Ex­plore user can spec­ify reac­tion or/​and gene knock outs in the ded­i­cated text field. To find the gene iden­ti­fiers, use the Gene Asso­ci­a­tion col­umn of the FBA tun­ing table. The reac­tions must be spec­i­fied with their iden­ti­fiers and not with their names. You can also select some reac­tions in the FBA tun­ing table and click on “Fill with selected reac­tions” to fill the text field with the iden­ti­fiers of the selected reac­tions.
Be care­ful, only the selected reac­tions that are dis­played will be filled. Use the FBA tun­ing table pager to be sure that all the selected reac­tions are displayed.

Select gene and reac­tions to knock out before com­put­ing the opti­mal objec­tive func­tion value

Flux Vari­abil­ity Analysis



Tuto­r­ial: how to com­pute FVA with MetExplore?

Flux Vari­abil­ity Analy­sis (FVA) is used to find the min­i­mum and max­i­mum flux for reac­tions in the net­work while main­tain­ing the opti­mal objec­tive func­tion value.
Since the FVA for all the reac­tions of a genome-​scale model can quite long (around 10 min­utes), it’s use­ful to spec­ify only some reac­tions for which the FVA will be per­formed. Note that the whole genome is used to com­pute FVA, even if only some reac­tions are selected. This means that the min­i­mum and max­i­mum fluxes will be com­puted only for the selected reac­tions.
The reac­tions must be spec­i­fied with their iden­ti­fiers and not with their names. You can also select some reac­tions in the FBA tun­ing table and click on “Fill with selected reac­tions” to fill the text field with the iden­ti­fiers of the selected reac­tions.
Be care­ful, only the selected reac­tions that are dis­played will be filled. Use the FBA tun­ing table pager to be sure that all the selected reac­tions are displayed.

Selec­tion of the reac­tions for which the FVA will be computed

After click­ing on “Sub­mit”, a result table is dis­played where appear the min­i­mum and the max­i­mum flux for each reac­tion that allow the objec­tive func­tion value to be optimal.

FVA Result Table

Below the table, the link “Result table” allows to down­lad the table as a tab­u­lated file directly read­able in any text edi­tor.

When the num­ber of selected reac­tions is less than 20, a plot cor­re­spond­ing to the table is dis­played. For each reac­tion, a red point indi­cates the min­i­mum value and a green point the max­i­mum value.

FVA Plot

At last, it’s pos­si­ble to map the flux obtained by FVA by launch­ing directly Cytoscape from Met­Ex­plore.
Note: please ver­ify that you have the java vir­tual machine installed in your com­puter to be able to launch Cytoscape. You don’t need to install Cytoscape. Met­Ex­plore allows to use a pre­vi­ous Cytoscape ses­sion to map the FVA results. This is par­tic­u­larly inter­est­ing if you already tuned the lay­out of meta­bolic networks.

Launch Cytoscape directly from Met­Ex­plore to map FVA results or map them on exist­ing Cytoscape sessions

When you click on “Sub­mit”, a new page is dis­played with the link to launch Cytoscape and the attribute files that can be read into a text edi­tor or loaded in other Cytoscape sessions.

Cytoscape launcher and links to the attribute files used by Cytoscape

A Cytoscape attribute called “flux” is cre­ated by Met­Ex­plore. The aim with this attribute is to eas­ily iden­tify reac­tions that are always active in one or the other direc­tion. The flux attribute equals to the min­i­mum flux if both fluxes are pos­i­tive, to the max­i­mum flux if both fluxes are neg­a­tive, or it equals to 0 if min and max have dif­fer­ent sign. In Cytoscape, select the “flux” visual style, and visual prop­er­ties will be tuned to draw edges con­sid­er­ing the value of the attribute. A green edge means that the flux is always in the for­ward direc­tion. A red edge means that the flux is always in the back­ward direc­tion. The reversible fluxes are black and the null fluxes are trans­par­ent. The width of the edge reac­tions is dis­played pro­por­tion­naly to the flux attribute but the map­ping needs some­times to be newly per­formed by click­ing on vizMap­per and tun­ing the Edge Line Width attribute. The other visual prop­er­ties are the same than explained in the Cytoscape ses­sion. Please visit the Cytoscape offi­cial site to know more about the Cytoscape func­tion­al­i­ties.
The meta­bolic net­worki attrib­utes are also loaded as Cytoscape attrib­utes. You can thus browse nodes by path­ways or cre­ate new visual prop­er­ties to rep­re­sent min­i­mum or max­i­mum fluxes.

Met­Ex­plore FVA results mapped on selected reac­tions by Cytoscape



Flux Vari­abil­ity Com­par­i­son between two conditions

How to com­pare flux vari­abil­ity between two con­di­tions with MetExplore?

In Met­Ex­plore, it’s pos­si­ble to com­pare the FVA com­puted on two flux con­di­tions (i.e two sets of con­straints). In the same way than sim­ple FVA, it is pos­si­ble to deal only with a set of selected reac­tions or with the whole network.

Met­Ex­plore FVA com­par­i­son between two flux conditions

Click­ing on “Dis­play sec­ond con­di­tion para­me­ters” dis­plays a new para­me­ters table where the Met­Ex­plore user can spec­ify an other objec­tive value and other flux con­straints. Once the sec­ond set of para­mets has been filled in, click on “Sub­mit”. A new table is displayed.

Met­Ex­plore FVA com­par­i­son between two flux conditions

For each reac­tion are dis­played the min­i­mum and the max­i­mum flux val­ues that allow the objec­tive func­tion value to be opti­mal in each one of the two con­di­tions. A glyphe allows a quick inter­pre­ta­tion of the dif­fer­ences between the two flux con­di­tions. By mov­ing the mouse over a glyphe, its mean­ing is dis­played. By click­ing on the glyphe, a table con­tain­ing all the glyphes and their leg­end is dis­played.

Pic­tureFlux bound changes in the sec­ond condition
The flux bounds are identical.
Iden­ti­cal max­i­mum fluxes. Higher min­i­mum flux.
Iden­ti­cal max­i­mum fluxes. Lower min­i­mum flux.
Iden­ti­cal min­i­mum fluxes. Lower max­i­mum flux.
Iden­ti­cal min­i­mum fluxes. Higher max­i­mum flux.
Higher min­i­mum and max­i­mum fluxes.
Lower min­i­mum and max­i­mum fluxes.
Higher max­i­mum flux and lower min­i­mum flux.
Higher min­i­mum flux and lower max­i­mum flux.
The flux con­strained in one direc­tion in the first con­di­tion is con­strained in the other direc­tion in the sec­ond condition.
The flux con­strained in one direc­tion in the first con­di­tion can be found in the two direc­tions in the sec­ond conditions.
The flux is poten­tially active in the first con­di­tion and always null in the sec­ond condition.
The flux can be found in both direc­tions in the first con­di­tion and is con­strained in one direc­tion in the sec­ond condition.
The flux always null in the first con­di­tion can be active in the sec­ond condition

Glyphes used to inter­pret the Met­Ex­plore FVA com­par­i­son between two flux conditions

The link “Result Table” below enables the user to down­load the result table in a tab­u­lated file.

Vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on a reac­tion flux



How to com­pute the vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on a reac­tion flux with MetExplore?

In this Met­Ex­plore func­tion, the flux through one reac­tion is var­ied and the opti­mal objec­tive value is cal­cu­lated as a func­tion of this flux and the result is plot­ted. This reveals how sen­si­tive the objec­tive func­tion value is to a par­tic­u­lar reaction.

Para­me­ters for com­put­ing vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on a reac­tion flux

First, spec­ify the objec­tive func­tion in the flux para­me­ter table.

The para­me­ters, sep­a­rated by white spaces, are:

  • The iden­ti­fier of the reac­tion for which the flux will vary
  • The first flux value of the plot
  • the final value of the plot
  • The value of the step

For instance, the para­me­ters “R_​EX_​glc_​e_​–5 0 0.1″ means that we will vary the flux of the reac­tion “R_​EX_​glc_​e_​” from –5 to 0 by iter­a­tively incre­ment­ing by 0.1. The val­ues of the flux will be : –5, –4.9, –4.8, –4.7, … ‚…, –0.2, –0.1, 0. Be care­ful, be smart when you choose the value of the step not to cre­ate a so huge num­ber of points that Met­Ex­plore won’t be able to solve the prob­lem in a human-​scale time!

After sub­mit­ting, a plot and a table are displayed.

Vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on a reac­tion flux in MetExplore

Vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on two reac­tion fluxes

How to com­pute the vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on two reac­tion fluxes with MetExplore?

This time, the fluxes of two reac­tions are vary­ing and the effect on the opti­mal func­tion value is computed.

Para­me­ters for com­put­ing vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on two reac­tion fluxes

The way to spec­ify the para­me­ters is the same than for vary­ing one flux: add the para­me­ters for the sec­ond reac­tion just after the para­me­ters of the first reac­tion and sep­a­rate them by a white space (see Fig­ure above). Be care­ful, be smart when you choose the value of the step not to cre­ate a so huge num­ber of points that Met­Ex­plore won’t be able to solve the prob­lem in a human-​scale time!

After sub­mit­ting, a heat map and a table are displayed.

Vari­a­tion of the opti­mal objec­tive func­tion value depend­ing on two reac­tion fluxes



Detect­ing live reactions

Iden­ti­fi­ca­tion of live reac­tions, i.e those able to carry a steady state flux with MetExplore.

Live reac­tions are those able to carry a steady state flux con­sid­er­ing the flux con­straints indi­cated in the flux para­me­ter table and the steady state con­di­tion.
This does not depend on the objec­tive function.

Com­put­ing live reac­tions in MetExplore

After option­ally select­ing a set of reac­tions to test and sub­mit­ting, a table is dis­played with the live reac­tions iden­ti­fied among the selected reactions.

Detect­ing essen­tial reactions

Iden­ti­fi­ca­tion of essen­tial reactions

Reac­tions are con­sid­ered as essen­tial when their inac­ti­va­tion (by mak­ing null their lower and upper flux bounds) makes null the objec­tive func­tion value.

Com­put­ing essen­tial reac­tions in MetExplore

After option­ally select­ing a set of reac­tions to test and sub­mit­ting, a table is dis­played with the essen­tial reac­tions iden­ti­fied among the selected reactions.

Orphan metabo­lites

Iden­ti­fi­ca­tion of orphan metabolites

Orphan metabo­lites rep­re­sent “dead ends” in the meta­bolic net­work. They are inter­nal metabo­lites that are not pro­duced or not con­sumed, what pre­vent them to be used in any flux dis­tri­b­u­tion because they can not be bal­anced. The iden­ti­fi­ca­tion of orphan metabo­lites is par­tic­u­larly use­ful in the meta­bolic net­work curation.

Iden­ti­fy­ing orphan metabo­lites in MetExplore

After sub­mit­ting, the list of orphan metabo­lites is dis­played. Note that this iden­ti­fi­ca­tion does not depend on the objec­tive function.




EXPORT

Top

Down­load and Visualise

The Met­Ex­plore func­tion “Down­load and Visu­alise” allows to export a fil­tered meta­bolic net­work in sev­eral formats:

  • Clas­si­cal SBML for­mat: the xml for­mat ded­i­cated to the meta­bolic net­works. The SBML for­mat gen­er­ated by Met­Ex­plore is described here.
  • Extended SBML for­mat: a sbml for­mat with addi­tional speci­fici­ties proper to Met­Ex­plore (spec­i­fi­ca­tions here).
  • Graph for­mats: avail­able in Cytoscape for­mat (sif)
    1. Bipar­tite graphs.
    2. Com­pound graphs.
    3. Reac­tion graphs.

The meta­bolic graphs are described in the sec­tion above. If the BioSource is based on the BioCyc-​like data­base, the but­ton “Remove com­part­men­tal­iza­tion” allows the user to remove the com­part­men­tal­iza­tion infor­ma­tion when export­ing the meta­bolic net­work. This means that each com­pound will be rep­re­sented only one instead of dupli­cated for each com­part­ment where it has been local­ized. When the Met­Ex­plore user selects a graph for­mat, the iden­ti­fiers are encoded in SBML for­mat (see here the SBML encod­ing). An option enables to deter­mine if self-​loops are allowed in the build­ing of the graph. A self-​loop is an edge that has the same node as source and as target.

Met­Ex­plore: down­load and visualise



Once an organ­ism selected and the meta­bolic net­work fil­tered (read the Fil­ters sec­tion), click on “sub­mit”. After a few sec­onds, the links to net­work, attribute files and to the Cytoscape launcher are displayed.

Met­Ex­plore: down­load and visualise



The gen­eral attrib­utes table con­tained in the file “Attrib­utes table” are described here.
When the Met­Ex­plore user launches Cytoscape from this page, the fil­tered meta­bolic net­work and the attrib­utes are directly loaded. It is thus pos­si­ble to adapt the visu­al­i­sa­tion style of Cytoscape to high­light infor­ma­tion about the meta­bolic net­work. Please refer to the Cytoscape doc­u­men­ta­tion about visual styles to know how to per­form visu­al­i­sa­tion tunings.



Top

Cre­ate Graph From SBML

This Met­Ex­plore func­tion allows the user to upload his (her) own SBML file and to export it into two graph for­mats described in the sec­tion above : com­pound graph and reac­tion graph. SBML file has to be com­pat­i­ble with the SBML speci­fici­ties. If you are not sure about the for­mat of your file, please use the online SBML val­ida­tor to check its syntax.

Met­Ex­plore: export SBML file in graphs



The Met­Ex­plore user can encode the iden­ti­fiers in SBML for­mat (see here the SBML encod­ing). An option enables to deter­mine if self-​loops are allowed in the build­ing of the graph. A self-​loop is an edge that has as source and tar­get the same node.

After click­ing on “Launch”, a link to the graph cre­ated from the SBML file and the Cytoscape launcher to directly visu­alise it are displayed.

Met­Ex­plore: export SBML file in graphs



Top

SBML For­mat

The SBML for­mat gen­er­ated by Met­Ex­plore has been inspired by the SBML for­mat used at the Sys­tems Biol­ogy Research Group of the UCSD .
Visit the SBML Site to see the global descrip­tion of SBML files.

  1. The iden­ti­fiers of the com­pounds (species) and the reac­tions from the path­way tools are trans­formed to be com­pat­i­ble with the SBML for­mat in the same way that the LISP func­tion “biocyc2sbml.lisp” devel­oped by Jeremy Zucker.

    • SBML unique iden­ti­fiers may only con­tain num­bers, let­ters or under­scores. Fur­ther­more, the first char­ac­ter can­not be a num­ber. Accord­ing to the BNF gram­mar on page 7 of the level 2 SBML spec:

      let­ter ::= ‘a’..‘z’, ‘A’..‘Z’
      digit ::= ‘0’..‘9’ nameChar ::= let­ter | digit | ‘_​’ name ::= ( let­ter | ‘_​’ ) nameChar*

      This is in con­trast to Bio­cyc unique iden­ti­fiers which may con­tain paren­the­ses, dashes, or html markup.

      In order to ensure that no infor­ma­tion is lost when con­vert­ing a Bio­cyc id to an SID, the fol­low­ing algo­rithm is employed:

      1. If the first char­ac­ter is not a let­ter, prepend a sin­gle under­score. i.e. 2-​OCTAPRENYLPHENOL becomes _​2-​OCTAPRENYLPHENOL

      2. For each char­ac­ter in the Bio­cyc id, if the char­ac­ter is not alphanu­meric or under­score, replace the char­ac­ter with its ascii value delim­ited by a dou­ble under­score. i.e. _​2-​OCTAPRENYLPHENOL becomes _​2_​_​45_​_​OCTAPRENYLPHENOL

      Note that this algo­rithm is reversible as long as Bio­cyc never uses an under­score at the begin­ning of an id and never hap­pens to have an id with a num­ber delim­ited by dou­ble under­scores. For­tu­nately, it does not.
    • In the notes sec­tion, XHTML does not appear to rec­og­nize enti­ties such as β and γ. Thus, these strings are replaced :

      1. & becomes &
      2. < becomes &lt;
      3. > becomes &gt;
      4. ” becomes &quot;
      5. &apos; becomes &apos;
    • Coef­fi­cients of a reac­tion had to be nor­mal­ized in order to be accepted by the SBML spec. For­tu­nately, the newest level 2 spec­i­fi­ca­tion accepts float­ing point num­bers for sto­i­chiom­e­try:

      Bio­cyc coef­fi­cient ==> SBML sto­i­chiom­e­try
      N ==> 1

      2N ==> 2

      M ==> 1

      0.5d0 ==> 0.5

  2. The rela­tion­ships between genes to enzymes to reac­tions are writ­ten in the notes tag.
    Exam­ple :

        <notes>
        <html:p>GENE_ASSOCIATION: ( BU278_trpB ) or ( BU277_trpA )</html:p>
        <html:p>PROTEIN_ASSOCIATION: ( Tryptophan synthase beta chain//RXN0-2382//TRYPSYN-RXN//Tryptophan synthase ) or ( Tryptophan synthase alpha chain//TRYPSYN-RXN//Tryptophan synthase )</html:p>    ...
        </notes>
        

    This indi­cates that this reac­tion can be catal­ysed by two dif­fer­ent enzymes (sur­rounded with brack­ets and sep­a­rated by “or”), each one com­posed by sev­eral monomers (sep­a­rated by “and”) coded by one gene. Be care­ful, this nota­tion does not take into account the splic­ing genes.
    NA” indi­cates that no infor­ma­tion is avail­able about the gene or the protein.

  3. The meta­bolic path­ways where occurs the reac­tion are indi­cated in the “notes tag” in the “<html:p>SUBSYSTEM:” field.
    Exam­ple :

        <notes>
        ...
        <html:p>SUBSYSTEM: glutathione biosynthesis</html:p>
        <html:p>SUBSYSTEM: &gamma;-glutamyl cycle</html:p>
    
        ...
        </notes>
        

    NA” indi­cates that the reac­tion does not occur in any meta­bolic pathway.

  4. The EC Num­ber is indi­cated in the “notes tag” in the “<html:p>PROTEIN_CLASS” tag.
    Exam­ple :

        <notes>
        ...
        <html:p>PROTEIN_CLASS: 6.3.2.2</html:p>
        ...
        </notes>
        

    NA” indi­cates that the infor­ma­tion is not available.
  5. When they are avail­able, the side com­pounds indi­cated by Bio­Cyc are indi­cated for each reac­tion. A com­pound is indi­cated as side com­pound in a reac­tion if Bio­Cyc rep­re­sents it as side com­pound in each meta­bolic path­way where the reac­tion is involved. Each side-​compound is indi­cated in the “notes”.
    Exam­ple:

        <notes>
        ...
        <html:p>SIDE: ADP</html:p>
        <html:p>SIDE: ATP</html:p>
        ...
        </notes>
        
  6. When the infor­ma­tion is avail­able, some com­pounds are indi­cated as cofac­tors in a reac­tion. The list of cofac­tor trans­for­ma­tions used to mark the com­pounds are avail­able here. Exam­ple:

        <notes>
        ...
        <html:p>COFACTOR: ADP</html:p>
        <html:p>COFACTOR: ATP</html:p>
        ...
        </notes>
        
  7. Also in the notes, the term ‘generic’ indi­cates if the reac­tion involve class com­pounds.
    Exam­ple:

        <notes>
        ...
        <html:p>GENERIC: false</html:p>
        ...
        </notes>
        
  8. In Bio­Cyc, the reac­tions are clas­si­fied in small-​molecule or macro­mol­e­cule reac­tions. When this infor­ma­tion is avail­able, the term ‘type’ indi­cates if the reac­tion is clas­si­fied in either clas­si­fi­ca­tion.
    Exam­ple:

        <notes>
        ...
        <html:p>TYPE: small</html:p>
        ...
        </notes>
        or
        <notes>
        ...
        <html:p>TYPE: macro</html:p>
        ...
        </notes>
        
  9. If the source data­base, the reac­tion is not assigned as spon­ta­neous and no enzyme has been assigned to catal­yse it, the reac­tion is con­sid­ered as a ‘hole’. In the Met­Ex­plore sbml for­mat, this infor­ma­tion, when avail­able, is indi­cated in the notes.
    Exam­ple:

        <notes>
        ...
        <html:p>HOLE: true</html:p>
        ...
        </notes>
        

Top

Met­Ex­plore SBML Format

This for­mat is used to com­plete the infor­ma­tions con­tained in a clas­si­cal sbml for­mat. As it is not com­pat­i­ble with the sbml spec­i­fi­ca­tions, this for­mat could not be used in main sbml-​using soft­wares. How­ever, this for­mat can be used in other appli­ca­tions and per­mits to get addi­tional infor­ma­tions.
The reac­tion and species iden­ti­fiers have been trans­formed to be sbml com­pat­i­ble, as indi­cated above. Other iden­ti­fiers (gene, pro­tein, path­way) have not been trans­formed.

The main dif­fer­ences with the clas­si­cal sbml for­mat are :

  1. The species ele­ment con­tains three new attributes:

    • mass: the mol­e­c­u­lar weight of the compound

    • for­mula: the mol­e­c­u­lar for­mula of the compound

    • generic: if generic=true, indi­cates that this com­pound reprensents a class of com­pounds and not an indi­vid­ual compound

  2. The reac­tion ele­ment con­tains 3 new attrib­utes and 5 new child nodes:

    • The EC num­ber, when avail­able, is indi­cated in the ‘ec’ attribute.

    • If in the source data­base, the reac­tion is not assigned as spon­ta­neous and no enzyme has been assigned to catal­yse it, the reac­tion is con­sid­ered as a ‘hole’. In the Met­Ex­plore sbml for­mat, this infor­ma­tion, when avail­able, is indi­cated in the hole attribute in the reac­tion tag.

    • The attribute ‘generic’ indi­cates if the reac­tion involve class compounds

    • The gene-​protein-​reaction (GPR) link is indi­cated thanks to enzyme ele­ments which con­tain pro­tein and gene ele­ments.
      Exam­ple:

      <reaction id="KETOLACTOSE__45__RXN" name="&beta;-galactosidase" reversible="false">
          <enzyme id="ENSG00000163521-MONOMER" name="beta-galactosidase precursor">
            <protein id="ENSG00000163521-MONOMER" name="beta-galactosidase precursor">
              <gene id="ENSG00000163521" name="MGC10771"/>
            </protein>
          <enzyme>
          <enzyme id="ENSG00000166103-MONOMER" name="beta-galactosidase precursor">
            <protein id="ENSG00000166103-MONOMER" name="beta-galactosidase precursor">
              <gene id="ENSG00000166103" name="NA"/>
            <protein>
          </enzyme>
          ...
    • Each path­way where the reac­tion is involved is indi­cated in the path­way tag. Example:

      <reaction id="ORNITHINE__45__GLU__45__AMINOTRANSFORASE__45__RXN" name="Ornithine--oxo-acid aminotransferase" reversible="true" hole="true" generic="false" type="small">
      ...
        <pathway id="PWY-4981" name="proline biosynthesis V (from arginine)"/>
        <pathway id="ARGININE-SYN4-PWY" name="arginine biosynthesis IV"/>
        <pathway id="PWY-3341" name="proline biosynthesis III"/>
        <pathway id="PWY-5004" name="superpathway of citrulline metabolism"/>
        <pathway id="ARGININE-DEG1-PWY" name="arginine degradation VII"/>
      ...
    • When they are avail­able, the side com­pounds indi­cated by Bio­Cyc are indi­cated for each reac­tion. A com­pound is indi­cated as side com­pound in a reac­tion if Bio­Cyc rep­re­sents it as side com­pound in each meta­bolic path­way where the reac­tion is involved. Each side-​compound is indi­cated in the “sides” tag.
      Exam­ple:

      <reaction id="ORNITHINE__45__GLU__45__AMINOTRANSFORASE__45__RXN" name="Ornithine--oxo-acid aminotransferase" reversible="true" hole="true" generic="false" type="small">
          ...
          <side-compounds>
            <speciesReference species="ADP" />
            <speciesReference species="ATP" />
          </side-compounds>
          ...
          
    • When the infor­ma­tion is avail­able, some com­pounds are indi­cated as cofac­tors in a reac­tion. The list of cofac­tor trans­for­ma­tions used to mark the com­pounds are avail­able here. Exam­ple:

         <reaction id="ORNITHINE__45__GLU__45__AMINOTRANSFORASE__45__RXN" name="Ornithine--oxo-acid aminotransferase" reversible="true" hole="true" generic="false" type="small">
          ...
          <cofactors>
            <speciesReference species="ADP" />
            <speciesReference species="ATP" />
          </cofactors>
          ...
          



Top

Meta­bolic graphs in MetExplore

The dif­fer­ent meta­bolic graph reprensen­ta­tions avail­able in Met­Ex­plore are dis­played in the Fig­ure above.

Met­Ex­plore: export meta­bolic net­works in graphs



In the com­pound graph, nodes cor­re­spond to metabo­lites and there is an edge between two metabo­lites if it exists a reac­tion where one is the sub­strate and the other the prod­uct.
In the reac­tion graph, nodes cor­re­spond to reac­tions. There is an edge between two reac­tions if one pro­duces a metabo­lite that is con­sumed by the other one.
In a bipar­tite graph the set of nodes can be divided in two sub­sets. One kind of node can only be con­nected to another kind of node. In the case of meta­bolic net­works one set of nodes cor­re­sponds to the metabo­lites and the other one to the reactions.



COM­MON FEATURES

Top

Met­Ex­plore FILTERS

Met­Ex­plore enables to fil­ter a meta­bolic net­work in sev­eral ways before apply­ing any Met­Ex­plore func­tion to restrict the scope of the study or to avoid sources of mis­in­ter­pre­ta­tions in meta­bolic graph analy­sis. Each fil­ter has a default value. Once a meta­bolic net­work selected, click on the link “Dis­play advanced fil­ters” to tune each Met­Ex­plore filter.

Gen­eral filters

These fil­ters affect the whole meta­bolic net­work. They are avail­able only for BioCyc_​like data­bases stored in MetExplore

Met­Ex­plore gen­eral filters



  • Remove path­way Holes. By default, this fil­ter is acti­vated. Path­way holes are reac­tions that are involved in the path­ways iden­ti­fied in the selected organ­ism but for which no enzymes are assigned.

  • Remove big mol­e­cule reac­tions. By default, this fil­ter is acti­vated. This fil­ter removes all the reac­tions that involve big mol­e­cules, such as reac­tions involv­ing nucleic acids or pro­teins. As we cur­rently deal only with BioCyc-​like data­bases, we use the Bio­Cyc ontol­ogy to clas­sify reac­tions between small-​molecule reac­tions and macro­mol­e­cule reactions.

  • Remove the generic reac­tions. By default, this fil­ter is inac­ti­vated. This fil­ter removes all the reac­tions involv­ing generic com­pounds (e.g. “an aldehyde”).

  • Get only pathway-​reactions By default, this fil­ter is inac­ti­vated. This fil­ter removes all the reac­tions not clas­si­fied into a meta­bolic path­way iden­ti­fied in the selected organism.

  • Remove side com­pounds. By default, this fil­ter is inac­ti­vated. As we cur­rently deal only with BioCyc-​like data­bases, side com­pounds are iden­ti­fied thanks to the infor­ma­tion about meta­bolic path­ways stored into the Bio­Cyc ontol­ogy. Some com­pounds are con­sid­ered as pri­mary com­pounds as they are involved in the back­bone of the meta­bolic path­way. The other metabo­lites are con­sid­ered as sec­ondary com­pounds. Most often, the sec­ondary com­pounds cor­re­spond to cofac­tors as ATP or NADH. In Met­Ex­plore, a com­pound is con­sid­ered as a sec­ondary com­pound of a reac­tion if it is anno­tated as sec­ondary com­pound in each meta­bolic path­way where the reac­tion is involved. When the fil­ter is acti­vated, the sec­ondary com­pounds of each reac­tion are removed.

  • Remove pairs of cofac­tors. By default, this fil­ter is acti­vated. This fil­ter uses another strat­egy to deal with cofac­tors. We estab­lished a list of 62 trans­for­ma­tions of cofac­tors shown in the table below. For instance, if a part of a reac­tion implies the fol­low­ing trans­for­ma­tion (or the reverse) : ATPADP + Pi, these three metabo­lites are removed in the reac­tion if the fil­ter is acti­vated. On the con­trary of the pre­vi­ous fil­ter, this one does not use path­way infor­ma­tion and then deal also with reac­tions not clas­si­fied in any meta­bolic pathway.

Names of the left compounds

Names of the right compounds

L-​glutamate

Glu­t­a­mine

2-​oxoglutarate

L-​glutamate

an oxi­dized elec­tron acceptor

a reduced elec­tron acceptor

S-​adenosyl-​L-​homocysteine

S-​adenosyl-​L-​methionine

ADP + phos­phate

ATP

adenosine-5′-phosphate + diphos­phate

ATP

a quinone

a hydro­quinone

dihy­dro­biopterin

tetrahy­dro­biopterin

CDP + phos­phate

CTP

CMP + diphos­phate

CTP

CMP

CMP-​N-​acetylneuraminate

coen­zyme A

acetyl-​CoA

coen­zyme A

an acyl-​CoA

coen­zyme A

benzoyl-​S-​CoA

coen­zyme A

caffeoyl-​CoA

coen­zyme A

4-​coumaroyl-​CoA

coen­zyme A

pimeloyl-​CoA

coen­zyme A

a fatty acyl CoA

coen­zyme A

malonyl-​CoA

coen­zyme A

®-methylmalonyl-​CoA

coen­zyme A

palmitoyl-​CoA

coen­zyme A

propionyl-​CoA

coen­zyme A

succinyl-​CoA

an oxi­dized cytochrome c

a reduced cytochrome c

an oxi­dized cytochrome c

a reduced cytochrome c

2′-deoxyadenosine-5′-diphosphate + phos­phate

dATP

a deoxynu­cle­o­side

a 2′-deoxyribonucleoside monophosphate

7,8-dihydrofolate

5,10-methylene-THF

dithio­thre­itol

oxi­dized dithiothreitol

a reduced elec­tron acceptor

an oxi­dized elec­tron acceptor

reduced elec­tron trans­fer­ring flavoprotein

oxi­dized elec­tron trans­fer­ring flavoprotein

FAD

FADH2

FMN

FMNH2

guanosine-5′-diphosphate

GDP-​α-​D-​mannose

guanosine-5′-diphosphate + phos­phate

guano­sine 5′-triphosphate

IDP

ITP

NAD℗H

NAD℗+ + H+

NADPH

NADP+ + H+

NAD+ + H+

NADH

oxaloac­etate

L-​aspartate

an oxi­dized thioredoxin

a reduced thioredoxin

adenosine-3′,5′-bisphosphate

phosphoadenosine-5′-phosphosulfate

pyrrolo­quino­line quinone

reduced pyrroloquinoline-​quinone

a pro­tein dithiol

a pro­tein disulfide

a pro­tein histidine

a pro­tein Nτ-​methyl-​histidine

pyru­vate

L-​alanine

a reduced adrenal ferredoxin

an oxi­dized adrenal ferredoxin

reduced cofac­tor F420

cofac­tor F420

a reduced ferredoxin

an oxi­dized ferredoxin

a reduced flavoprotein

an oxi­dized flavoprotein

a reduced rubredoxin

an oxi­dized rubredoxin

suc­ci­nate + CO2 + H2O

2-​oxoglutarate + oxy­gen

dTDP

TDP-​rhamnose

tetrahy­dro­fo­late

10-​formyl-​tetrahydrofolate

tetrahy­dro­fo­late

5,10-methylene-THF

a thiore­doxin

an oxi­dized thioredoxin

a ubiquinol

a ubiquinone

uridine-5′-diphosphate + phos­phate

UTP

uridine-5′-diphosphate

UDP-​D-​galactose

uridine-5′-diphosphate

UDP-​D-​glucose

uridine-5′-diphosphate

UDP-​D-​glucuronate

uridine-5′-diphosphate

UDP-​L-​rhamnose

uridine-5′-diphosphate

UDP-​N-​acetyl-​D-​glucosamine

uridine-5′-diphosphate

UDP-​N-​acetyl-​D-​galactosamine

Sulfurated-​Sulfur-​Acceptors

Unsulfurated-​Sulfur-​Acceptors

Table of trans­for­ma­tions used in Met­Ex­plore to fil­ter the cofac­tors in the reactions



Spe­cific filters

The Met­Ex­plore user can spec­ify only some meta­bolic path­ways to keep or some metabo­lites to remove before apply­ing the Met­Ex­plore functions.

  • Select path­ways to keep. The image below shows the Met­Ex­plore fil­ter to keep or remove path­ways to restrict the scope of the analy­sis. The first col­umn con­tains the name of the path­ways iden­ti­fied in the selected organ­ism. Each path­way is linked to the cor­re­spond­ing page in the cor­re­spond­ing source data­base (pass­words to the data­base access are indi­cated depend­ing on the data­base). The sec­ond col­umn indi­cates the pro­por­tion of catal­ysed reac­tions in the meta­bolic path­ways. The third colum is a check box that indi­cates if the path­way will be kept in the analy­sis. If a meta­bolic path­way is not checked, the reac­tions that it involves are removed from the net­work if they don’t appear in any other checked meta­bolic path­way. Note: if you don’t want to keep reac­tions not clas­si­fied into a path­way, acti­vate the fil­ter “Get only pathway-​reactions” in the gen­eral filters.



Met­Ex­plore Path­way Filter



  • Select com­pounds to keep. The image below shows the Met­Ex­plore fil­ter to keep or remove some metabo­lites in the fil­tered meta­bolic net­work. The first col­umn con­tains all the com­pounds present in the meta­bolic net­work of the selected organ­ism. The sec­ond col­umn indi­cates the num­ber of reac­tions where each metabo­lite is involved. The third col­umn con­tains a check box that indi­cates if the metabo­lite will be kept in the meta­bolic net­work. On the con­trary to the gen­eral fil­ters that deal with cofac­tors, the com­pounds not checked here are com­pletely removed from the meta­bolic net­work. Eight com­pounds are by default removed: water, pro­ton, CO2, phos­phate, diphos­phate, NH3, H2O2 and O2, as they are com­monly involved as sec­ondary com­pounds in reac­tions. Note: if a metabo­lite is removed by the acti­va­tion of other fil­ters, check­ing it in this fil­ter has no effect.

Met­Ex­plore Com­pound Filter




Top

Topo­log­i­cal infor­ma­tion glyphes

In the results of the Met­Ex­plore func­tion, the topo­log­i­cal infor­ma­tion of the metabo­lites is dis­played in the form of glyphes. The table below shows the mean­ing of each glyph.

The metabo­lite is con­sumed and pro­duced by sev­eral reactions

Source — The metabo­lite is not pro­duced by any reac­tion and con­sumed by sev­eral reactions

Dead End — The metabo­lite is not con­sumed by any reac­tion and pro­duced by sev­eral reactions

Choke Point — The metabo­lite is uniquely con­sumed by a spe­cific reac­tion and pro­duced by sev­eral reactions

Choke Point — The metabo­lite is uniquely pro­duced by a spe­cific reac­tion and con­sumed by sev­eral reactions

Choke Point — The metabo­lite is uniquely pro­duced by a spe­cific reac­tion and uniquely pro­duced by an other reaction

Dead End Choke Point — The metabo­lite is uniquely pro­duced by a spe­cific reac­tion and not pro­duced by any reaction

Source Choke Point — The metabo­lite is uniquely con­sumed by a spe­cific reac­tion and not used by any reaction

Source or Dead End Choke Point — The metabo­lite is uniquely con­sumed or pro­duced by the same reversible reaction

Tables of glyphes rep­re­sent­ing topo­log­i­cal fea­tures of the metabo­lites in a fil­tered meta­bolic network.




Top

Using Cytoscape with MetExplore

Met­Ex­plore func­tions enable to directly load a fil­tered meta­bolic net­work and its attrib­utes into a ver­sion of Cytoscape tuned for Met­Ex­plore.

Require­ments to use Cytoscape with Met­Ex­plore
In Met­Ex­plore, Cytoscape is run via Java Web Start. Any browser that sup­ports the Java ver­sion 1.5 or higher can launch Cytoscape from Met­Ex­plore.

Visual style of the Met­Ex­plore meta­bolic net­works in Cytoscape
The visual style we use to draw the meta­bolic net­works loaded from Met­Ex­plore is an exten­sion of the SBML visual style present in Cytoscape. In par­tic­u­lar, the Met­Ex­plore visual style allows to rep­re­sent the reversible reac­tions.
The leg­end of the Met­Ex­plore visual style is dis­played in the table below.

Irre­versible reaction

Reversible reac­tion

Metabo­lite

Link from a metabo­lite to a reac­tion using it

Link from a metabo­lite to a reac­tion pro­duc­ing it



Gen­eral net­work attrib­utes
Gen­eral attrib­utes of fil­tered meta­bolic net­works are down­load­able or directly loaded in Cytoscape in most of the Met­Ex­plore func­tions. Each line of the gen­eral attrib­utes table cor­re­sponds to a reac­tion or to a metabo­lite. The columns cor­re­spond to the fol­low­ing attributes:

  • ID: the iden­ti­fier of the reac­tion or of the metabo­lite in SBML for­mat,

  • sbml name: the com­mon name of the reac­tion or of the metabolite,

  • mass: the mass of the metabo­lites (NA for the reac­tions or metabo­lites with­out assigned mass),

  • for­mula: the for­mula of the metabo­lites (NA for reac­tions or metabo­lites with­out assigned formula),

  • path­ways: the list of path­ways where the reac­tion or the metabo­lite is involved, each path­way is sep­a­rated by “_​_​”,

  • ec: the EC num­ber of the reac­tions (NA for metabo­lites or reac­tions with­out assigned EC number,

  • reversibil­ity: the reversibil­ity (true or false for reac­tions, NA for metabolites)

  • sbml type: the type of the node (“species” for metabo­lites, “reac­tion” for reactions).



The same attrib­utes are loaded in Cytoscape when it is launched from Cytoscape. How­ever, to use the gen­eral attrib­utes table in another Cytoscape ses­sion, please refer to the Cytoscape Doc­u­men­ta­tion.