Quick Start

A basic work-flow with influx_si is composed of the following steps:

  1. Create a MTF file set (Multiple TSV Files) describing your metabolic reactions and carbon transitions (.netw), experimental data (.miso) label input (.linp) and some non mandatory measurements options (.mflux, .mmet, .tvar, .cntsr, .opt). Let an example MTF set has a prefix mynetwork. The syntax rules for reactions will be more or less obvious to someone working on metabolism biochemistry. So, to go quickly, you can inspire from example files test/mtf/e_coli.netw and co. distributed with the influx_si software (run influx_s --copy_test to bring them to your current directory). You can also consult the help message from txt2ftbl -h for --mtf option.

Note

NA values (as “Non Available”) are admitted as measurements values where appropriate. The difference with the situation where measurements are simply omitted is that NA measurements are simulated and are present in the vectors simulated unscaled labeling measurements and simulated scaled labeling measurements in the result kvh file.

Note

In case of influx_i, label kinetics can be provided in .miso file using non-empty Time column. Empty cells in Value column are equivalent to NA.

  1. Set your current directory to the directory of mynetwork.* and run

    $ influx_s.py --prefix mynetwork
    

or

$ influx_i.py --prefix mynetwork

depending on stationary or instationary labeling context. We suppose here that directory of influx_si binaries is in the PATH variable.

An influx_si run will produce the following files in the same directory that mynetwok.*:

mynetwork.ftbl
FTBL is a previously used format as a front-end format. It is still in use but behind the scenes. This file can be necessary as entry for ftbl2* utilities.
mynetwork.log
contains the run-time output from various scripts, in particular, it contains a report on convergence history during the fitting process. It can be helpful for identifying potential problems, but if everything is going well, the user does not have to examine the content of this file;
mynetwork.err
contains the warning and error messages. Normally, this file should be empty (0 byte size);
mynetwork_res.kvh
contains all the results. KVH format is a lightweight plain text format for hierarchically structured data. It can be seen in a text editor or in a spreadsheet software as its fields are tab separated. It can also be processed by user’s custom software for post-processing, graphics output and alike. If influx_si is run on a series of starting points, there will be generated a common result file mynetwork_res.kvh which contains common information to all starting points but also a series of kvh files, one by starting point, e.g. mynetwork_res.V1.kvh, mynetwork_res.V2.kvh and so on;
mynetwork.pres.csv
contains a matrix of fitted parameters and final cost values. Each column corresponds to a particular starting point if run with --fseries and /or --iseries options. If influx_si was run without these options, the file will contain only one column corresponding to the starting point defined in the mynetwork.tvar file or to the random starting point.
edge.netflux.mynetwok, edge.xchflux.mynetwok, node.log2pool.mynetwork

as the middle name of these files suggest, they can be used to map the corresponding values on the network graph in the cytoscape software.

Note

All these files are silently overwritten if already exist. So take care to copy your results elsewhere if you want to protect them from overwriting.

custom files (e.g. mynetwork.pdf)
These files can be produced by user supplied scripts that are executed at the end of influx_si simulations. For example, we provide a script plot_ilab.R which can be used to plot label kinetics simulated by influx_i. One or many of such custom scripts can be given in .opt file, field posttreat_R (cf. e_coli_i.opt for example)

Note

It can be helpful to do some “dry runs” by executing

$ influx_s.py --noopt --pref mynetwork

before collecting actual measurement data to see if intended measurements will be sufficient to well define all fluxes, or at least the fluxes of interest. It is possible to do so because the measurement values in the .miso file have no impact on flux SD calculation when --noopt option is used. So it can be used any values, even NA at this moment. On the contrary, SD values set in .miso file, must be realistic. It is generally not a problem as they express measurements errors and are more or less known for a given measurement method.

It is worthwhile to stress that a “dry run” is done for some presumed free flux values. If they reveal to be very different from actual flux values, it can happen that a network considered as well defined at moment of “dry run” turned into a badly defined network with actual measurement data and corresponding estimated fluxes. So it is important to do his best to guess the most realistic free fluxes for “dry runs” and log their values in .tvar file.

  1. See warning and error messages in mynetwork.err if any. Correct what has to be corrected and retry p. 2
  2. Extract and use the numerical results from the mynetwork_res.kvh file.
  3. Optionally, visualize net fluxes (or exchange fluxes or logarithm of metabolite concentrations \(\log_2(M)\)) in cytoscape using ftbl2xgmml to produce a .xgmml file and then mapping edge.netflux.mynetwok.attrs, edge.xchflux.mynetwok.attrs or node.log2pool.mynetwork.attrs on graphical attributes like edge width, color etc. in cytoscape.