Master

analysis/microbiomeModelingToolbox/tutorial_mgPipe.m

@@ -109,7 +109,7 @@
 
 taxonLevel='Species';
 
-createPanModels(modPath,panPath,taxonLevel,numWorkers);
+createPanModels(modPath,panPath,taxonLevel);
 %% 
 % By setting panPath as the input variable modPath for initMgPipe, personalized 
 % microbiome models for the samples from the above study with the EMBL-EBI accession 
@@ -217,24 +217,26 @@
 % are computed:*
 % 
 % Flux Variability analysis for all the exchange reactions of the diet and fecal 
-% compartment are performed and saved in a file called "simRes" and exported in 
-% spreadsheet format in the following files:
-%% 
-% # *netSecretionFluxes.csv* contains the maximal secretion potential of each 
-% sample for each metabolite that could be produced by at least one microbe. Note 
-% that secretion potential is substracted by dietary uptake of each metabolite, 
-% hence, only the metabolite production explicitly resulting from microbiome metabolism 
-% are quantified and shown in the table. The unit is mmol/sample/day.
-% # *netUptakeFluxes.csv* contains the maximal uptake potential of each sample 
-% for each metabolite that could be taken up by at least one microbe. Note that 
-% uptake potential is substracted by the fraction of dietary metabolites that 
-% are only passing through and being excreted, hence, only the quantities of each 
-% metabolite that are actively being consumed by at least one microbe are shown 
-% in the table. The unit is mmol/sample/day.
-%% 
-% The similarity of metabolic secretion profiles (using the net secretion potential 
-% as features) between individuals is also evaluated with classical multidimensional 
-% scaling. 
+% compartment are performed and temporarily saved in a file called "simRes". Specifically 
+% what is temporarily saved is:
+%% 
+% # *fvaCt* a cell array containing min flux through uptake and max through 
+% secretion exchanges
+% # *nsCt* a cell array containing max flux through uptake and min trough secretion 
+% exchanges
+% # *presol* an array containing the value of objectives for each microbiota 
+% model with rich and selected diet
+% # *inFesMat* cell array containing the names of the microbiota models that 
+% reported an infeasible status when solved for their objective 
+%% 
+% Finally, the net uptake and secretion potential are computed in a metabolite 
+% resolved manner and saved in the output variables 'netSecretionFluxes' and netUptakeFluxes, 
+% and the files 'netSecretionFluxes.csv' and 'netUptakeFluxes.csv' in the results 
+% folder. They indicate the maximal uptake and production, respectively, of each 
+% metabolite and are computed as the absolute value of the sum of the maximal 
+% secretion flux with the maximal uptake flux. The similarity of metabolic secretion 
+% profiles (using the net secretion potential as features) between individuals 
+% is also evaluated with classical multidimensional scaling. 
 % 
 % The output file "ReactionAbundance.csv" in the Results folder contains the 
 % relative abundance of each reaction in each sample. A description of each reaction 
@@ -247,7 +249,7 @@
 % 
 % In the export of models with dietary constraints is desired, they will be 
 % found in the folder constrModelsPath.
-%% Correlation between computed uptake and secretion fluxes and abundances and different taxon levels
+%% Correlation between computed fluxes and abundances and different taxon levels
 % For an overview of metabolite-taxa relationships, the computed uptake and 
 % secretion profiles for each metabolite can be correlated with taxon abundances 
 % on different levels (species, genus, etc.). Note that a correlation may not 
@@ -267,12 +269,6 @@
 % Path to fluxes that should be correlated
 
 fluxPath = [pwd filesep 'Results' filesep 'inputDiet_net_secretion_fluxes.csv'];
-%% 
-% Correlations with uptake fluxes are computed similarly by using inputDiet_net_uptake_fluxes.csv 
-% as the input file instead.
-% 
-% Run the function.
-
 corrMethod = 'Spearman';
 [FluxCorrelations, PValues, TaxonomyInfo] = correlateFluxWithTaxonAbundance(abunFilePath, fluxPath, taxInfo, corrMethod);
 %% 
@@ -281,9 +277,8 @@
 % with annotations, e.g., in R.
 %% Metabolite-resolved plots of metabolite-strain correlations
 % For a more detailed view of the relationships between metabolites and abundances, 
-% the secretion fluxes can also be directly plotted against organism abundances. 
-% This may reveal microbes that are bottlenecks for the production of metabolites 
-% of interest.
+% the fluxes can also be directly plotted against organism abundances. This may 
+% reveal microbes that are bottlenecks for the production of metabolites of interest.
 % 
 % We will define the metabolites to analyze (default: all exchanged metabolites 
 % in the models). As an example, we will take only one metabolite, acetate. Enter 
@@ -301,10 +296,6 @@
 plotFluxesAgainstOrganismAbundances(abunFilePath,fluxPath,metList);
 %% 
 % Afterwards, the plots will be in the subfolder Metabolite_plots.
-% 
-% Note that the microbe-metabolite relationships for metabolite uptake can be 
-% plotted in the same way by using inputDiet_net_uptake_fluxes.csv as the input 
-% file instead.
 %% Stratification of samples
 % If metadata for the analyzed samples is available (e.g., disease state), the 
 % samples can be stratified based on this classification. To provide metadata, 
@@ -354,18 +345,17 @@
 % p-values and test results. If there are any fluxes or abundances that significantly 
 % differed between groups, there will be files ending in "significantFeatures.txt" 
 % listing only these instances. Created violin plots will be found in the folder 
-% "ViolinPlots". There will be an image in png format for each predicted metabolite's 
-% uptake and secretion potential.
-%% Targeted analysis: Strain-level contributions to the uptake and secretion of metabolites of interest
+% "ViolinPlots". There will be an image in png and pdf format for each predicted 
+% metabolite's uptake and secretion potential. There will also be a file ending 
+% in "All_plots.pdf" containing all plots.
+%% Targeted analysis: Strain-level contributions to metabolites of interest
 % For metabolites of particular interest (e.g., for which the community-wide 
 % secretion potential was significantly different between disease cases and controls), 
 % the strains consuming and secreting the metabolites in each sample may be computed. 
 % This will yield insight into the contribution of each strain to each metabolite. 
 % Note that for metabolites for which the community wide secretion potential did 
 % not differ, the strains contributing to metabolites may still be significantly 
-% different. The contribution of each strain to metabolite consumption is also 
-% computed through the same function. If pan-models are used, the contributions 
-% will be on the species, genus, etc. level.
+% different.
 % 
 % The first step for the preparation of targeted analyses is the export of models 
 % that had already been constrained with the simulated dietary regime. They can 
@@ -389,10 +379,7 @@
 % internal exchange reactions that had nonzero flux for each analyzed metabolite. 
 % The output 'maxFluxes' shows the corresponding forward fluxes. 'fluxSpans' shows 
 % the distance between minimal and maximal fluxes for each internal exchange reaction 
-% with nonzero flux for each metabolite. Note that contrary to conventions in 
-% the field, in this case, *minFluxes represent internal secretion fluxes* for 
-% each strain and each metabolite and *maxFluxes represent internal uptake fluxes* 
-% for each strain and each metabolite.
+% with nonzero flux for each metabolite.
 % 
 % Afterwards, statistical analysis of the strain contributions can also be performed.
 %