ecModel simulation tests

#ecModel-simulation-tests

This notebook contains tests to see if simulating the ecModel with proteomics works as intended in the simulations service of DeCaF.

Note: using the requirements from simulations (i.e. create a conda env and in it pip install -r ./requirements/requirements.txt)

Benjamín J. Sánchez, 2019-11-21

1. Loading Model and Data

#1.-Loading-Model-and-Data

2. Functions

#2.-Functions

3. Testing

#3.-Testing

Testing with made up data in E. coli:

#Testing-with-made-up-data-in-E.-coli:

Testing with real data in yeast:

#Testing-with-real-data-in-yeast:

Getting usage values and comparing with caffeine

#Getting-usage-values-and-comparing-with-caffeine
Loading output library...

4. Signs of Shadow Prices

#4.-Signs-of-Shadow-Prices

Let's evaluate what happens if we take the positive shadow prices (if any) instead of the negative ones, by sorting in a decreasing order instead of an increasing order:

No positive shadow prices are detected, and hence the algorithm gets stuck. Finally, let's take the absolute value of all shadow prices and then rank in decreasing order, so that the sign doesn't matter:

We see that we get the same results as with negative shadow prices, so no positive shadow prices are popping up in any of the flexibilization iterations.

Temporal, while the data upload interface at caffeine is still wip.