Recent changes list

Version 2.3.4

TR implementation, type C/c added, types D..F documented

Version 2.3.3

major bugfix in flux bounds, solutions not found

Version 2.3.2

modeset-score, new option --i, --W now documented
bug in remove-duplicates, apply-removables
Safe now if your bash is configured to alias rm='rm --i', similarly cp and mv
Fixed an issue by converting to numbers which were occasionally interpreted as strings and lead to wrong comparison results
FVA fixed, did not properly implement minimum
compute-FBA --m now minimizes in the sense that large negative values will be computed
compute-FBA --ma minimizes absolute values, as compute-FBA --m before
syntax of min/max items in simulations objective extended:
- an entry can be preceded by a number, weight in the objective function
- this number can even be negative
- several min max items can be combined
- min/max may not be the first identifier
- after min/max several entries may follow
- setting a flux can only done before the first min/max tag
- new minabs keyword: minimal absolute value

Version 2.3.1

modeset-score, specification of --W changed, explicit weights now applied after root computation
bug in simulations-work-assure when reaction-names is present
seqq extended, now simulate work with numbers
Parameters to the simulate function
fitness functions for negative targetfluxes fixed

Version 2.3.0

Scoring function rewritten:
- fluxmin --F/biomax --b/minintake --m/fitness --f/enzymecost --e can now be combined
- tradminintake --M removed
- fluxmin --F/biomax --b/minintake --m have now weight parameters
- enzyme cost minimization separated from flux minimization
- targetfluxes now also active for --b
Reimplementation of enzyme cost minimization: --e
modeset System, basic functions: allout2modeset modeset2allout
modeset-score
names_in_allout now also effects the concentrations section in allout.txt
new default for compute-FBA: no scoring function (just return a feasible solution)
--r and --delta parameters (compute-FBA) for relative flux calculations
Change of option names for Shlomi expression match (now --xs, was --x --xu --xl --xs)
simulate-noallout fixed
CPLEX options renamed: --e --> --CPLEXemph; --C --> --CPLEXline; --n --> --CPLEXnodes
option --max in compute-FBA, restricts absolute values of all fluxes, prevents unboundedness

Version 2.2.3

not anymore dependent on --enable-switch version of gawk
not anymore dependent on seq
installation section of manual and tutorial improved

Version 2.2.2

harmless Warning "concentration-ranges-molaric.fgf: No such file or directory" removed
major rewrite of reaction2fa
remove-duplicates/apply-removables functions
reactions-filterout/metabolites-filterout
if the file metabolites does not comply to file reactions it is simply repaired
character / removed in simulation names (will cause problems in allout2sbml etc.)
type m for fluxmin (number of nonzero fluxes)
illegal combinations of flux bounds (also different files combined) now are warned (e.g. 2 ... -1)
default bound for max and min optimizations raised from 1e4 to 1e6
FCA (Burgard et al.): make-FCA-simulationsfile/FCA-chart
Fitness type 8

Version 2.2.1

sbml2fa: better handling of empty compartments,names,ids
Bug in -f 7 fixed, wrong fitness values displayed
no more dependent on strings utility
Compatibility with CPLEX 12.2

Version 2.2.0

pure fitness maximization for -f without -F/-m Change of specification! For combination of fitness maximization and flux minimization now -f and -F must be given!
fitness issue for targetfluxes of zero fixed
fitness issue for negative targetfluxes fixed
new fitness types 4-7: 1-norm and max-norm/ w/o normalization
documentation for fluxmin-type and fitness types in 'compute-FBA -h'
Fitness written in evaluation.txt
Implementation of MOMA with the fitness types
Implementation of expression match score in compute-FBA (Huthmacher algo): see -x option
Efficiency improvements in cplex-std-call, FB-optimization-std-call
file: flux-weights for flux minimization (multiplied when also enzymes are used)
strict lower cap (1e-10) in equilibriums for TR removed, now caps are 1e-300...1e+300
ensure-standard-directions rewritten and documented in the manual
restrict-equilibriums rewritten, now accepts two forms and documented in the manual
equilibriums need not have an entry for all reactions anymore, 1 is the assumed default.
reactions-printout rewritten, reactions-cleartext-printout removed
number of fluxes not evaluator anymore, instead: 0=no solution, 1=any soulution
empty evaluator now allowed (+ if solution exists)
allout.txt creation rewritten, small bug (exchanges not at start) fixed, faster now
simulate-single-wrap now takes simulation description instead of name: much faster for large simulations file
extended syntax for concentration-class-ranges, see manual
prune-network: more verbose, now stores the solutions in allout.txt
zero modes/empty solutions fixed
print-sbml written, replaces make-sbml... functions
print-expa-network, print-metatool-network, print-sbml now omit FASIMU pseudo reactions
manual and tutorial reformatted and slightly rewritten, now in latex
bug in compute-FBA -T (log-equilibriums wrong)
prune-network now also generates modified fluxfix, fluxmin-weights, setfluxes, fluxbounds, expressions, enzymes

Version 2.1.7

small bug in simulationsfile x <= REA <= y syntax if x is negative
boundary condition in sbml2fa repaired
simulationsfile REA = y syntax
boundary syntax now also in files read in by the simulations file

Version 2.1.6

allout2eval (reevaluation of solutions)
various bugfixes
allout2sbml rewritten
if equilibriums file does not exist, trivial one is generated upon start of fabase

Version 2.1.5

constraintfluxes (undocumented) renamed to setfluxes
fluxbounds (file): additional layer of flux bounds
check-essentiality function
ranges in simulation constraints

Version 2.1.4

allout2netreactionfiles creates now an directory netreactions (similar to allout2bina, allout2valfiles)