Merge pull request #14 from MetExplore/development

Prepare release of v0.9.0-beta.2

README.md

@@ -1,21 +1,30 @@
-# Introduction
+<div align="center">
+    <a href="https://metexplore.github.io/miom"><img align="center" src="https://github.com/MetExplore/miom/raw/development/docs/assets/img/miom_v1.png" alt="MIOM" width="545" title="MIOM: Mixed Integer Optimization for Metabolism"/>
+    </a>
+</div>
+
+<div align="center">
+
 [![Try It Online](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1JAOEHLlRCW8GziIpBqkFwJL2ha3OEOWJ?usp=sharing)
-[![PyPI version](https://badge.fury.io/py/miom.svg)](https://badge.fury.io/py/miom)
+[![PyPI](https://img.shields.io/pypi/v/miom?style=flat-square&label=PyPI&logo=pypi&logoColor=white)](https://pypi.org/project/miom/)
 ![Tests](https://github.com/metexplore/miom/actions/workflows/unit-tests.yml/badge.svg)
 
+</div>
+
+---
+
 https://metexplore.github.io/miom
 
 __MIOM__ (Mixed Integer Optimization for Metabolism) is a python library for creating and solving complex optimization problems using genome-scale metabolic networks, in just a few lines. 
 
-MIOM offers a high-level API that leverages the power of modern Mixed Integer Optimization (MIO) solvers to easily define steady-state metabolic optimization problems, from simple Flux Balance Analysis (FBA) simulations, to more complex problems, such as sparse FBA or context-specific reconstruction problems, and solve them the __required level of optimality__.
-
-MIOM uses the [PICOS](https://picos-api.gitlab.io/picos/) and the [Python-MIP](https://www.python-mip.com/) libraries to build and solve the optimization problems using many commercial, academic and free solvers. It is also compatible and complementary to [cobrapy](https://opencobra.github.io/cobrapy/).
+MIOM offers a high-level API that leverages the power of modern Mixed Integer Optimization (MIO) solvers to easily define steady-state metabolic optimization problems, from simple Flux Balance Analysis (FBA) simulations, to more complex problems, such as sparse FBA or context-specific reconstruction algorithms, and solve them the __required level of optimality__. It supports many free and commercial solvers thanks to the integration with the [PICOS](https://picos-api.gitlab.io/picos/) and the [Python-MIP](https://www.python-mip.com/). It is also compatible and complementary to [cobrapy](https://opencobra.github.io/cobrapy/).
 
-Here is an example of how to implement FBA and Sparse FBA to maximize flux through the biomass reaction in the Recon3D model with MIOM:
 
 ```python
 import miom
 
+# Download the Recon3D metabolic network and find the maximum flux value
+# through the biomass_reaction
 model = (miom
         .load('@BiGG/Recon3D.miom')
         .steady_state()
@@ -25,7 +34,9 @@ model = (miom
 print("Optimal flux:", model.get_fluxes('biomass_reaction'))
 print("Number of active reactions:", sum(abs(model.get_fluxes()) > 1e-8))
 
-# Minimize the number of reactions preserving the optimal flux
+# Transform the previous FBA optimization problem into a Sparse FBA problem (MIP).
+# Solving this problem returns the minimum number of active reactions preserving
+# the optimal flux through the biomass_reaction
 V, X = (model
         .setup(opt_tol=0.05)
         .set_fluxes_for('biomass_reaction')
@@ -77,7 +88,7 @@ To make things even easier, the method `load_gem` can import any model from this
 
 ```python
 human1 = miom.mio.load_gem("@SysBioChalmers/Human-GEM.miom")
-recon3d = miom.mio.load_gem("@BiGG/Human-GEM.miom")
+recon3d = miom.mio.load_gem("@BiGG/Recon3D.miom")
 ```
 
 Here is an example of how to load a metabolic network and maximize the flux through a target reaction using FBA, and then how to modify the original problem to implement the sparse FBA problem adding only a few lines to the original problem:

docs/index.md

@@ -1,19 +1,24 @@
-# Introduction
+<a href="https://metexplore.github.io/miom"><img src="assets/img/miom_v1_fit.png" alt="MIOM" title="MIOM: Mixed Integer Optimization for Metabolism"/></a>
+
 [![Try It Online](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1JAOEHLlRCW8GziIpBqkFwJL2ha3OEOWJ?usp=sharing)
-[![PyPI version](https://badge.fury.io/py/miom.svg)](https://badge.fury.io/py/miom)
+[![PyPI](https://img.shields.io/pypi/v/miom?style=flat-square&label=PyPI&logo=pypi&logoColor=white)](https://pypi.org/project/miom/)
 ![Tests](https://github.com/metexplore/miom/actions/workflows/unit-tests.yml/badge.svg)
 
+# Introduction
+
 __MIOM__ (Mixed Integer Optimization for Metabolism) is a python library for creating and solving complex optimization problems using genome-scale metabolic networks, in just a few lines. 
 
-MIOM offers a high-level API that leverages the power of modern Mixed Integer Optimization (MIO) solvers to easily define steady-state metabolic optimization problems, from simple Flux Balance Analysis (FBA) simulations, to more complex problems, such as sparse FBA or context-specific reconstruction problems, and solve them the __required level of optimality__.
+MIOM offers a high-level API that leverages the power of modern Mixed Integer Optimization (MIO) solvers to easily define steady-state metabolic optimization problems, from simple Flux Balance Analysis (FBA) simulations, to more complex problems, such as [sparse FBA](https://metexplore.github.io/miom/examples/sparse_fba/) or context-specific reconstruction algorithms (see for example [iMAT](https://metexplore.github.io/miom/examples/imat/) or [Fastcore](https://metexplore.github.io/miom/examples/fastcore/)), and solve them the __required level of optimality__. It supports many free and commercial solvers thanks to the integration with the [PICOS](https://picos-api.gitlab.io/picos/) and the [Python-MIP](https://www.python-mip.com/). It is also compatible and complementary to [cobrapy](https://opencobra.github.io/cobrapy/).
 
-MIOM uses the [PICOS](https://picos-api.gitlab.io/picos/) and the [Python-MIP](https://www.python-mip.com/) libraries to build and solve the optimization problems using many commercial, academic and free solvers. It is also compatible and complementary to [cobrapy](https://opencobra.github.io/cobrapy/).
 
-Here is an example of how to implement FBA and Sparse FBA to maximize flux through the biomass reaction in the Recon3D model with MIOM:
+!!! warning
+    This library is functional but still in a very early stage. API is still not stable and might be changed in future versions.
 
 ```python
 import miom
 
+# Download the Recon3D metabolic network and find the maximum flux value
+# through the biomass_reaction
 model = (miom
         .load('@BiGG/Recon3D.miom')
         .steady_state()
@@ -23,7 +28,9 @@ model = (miom
 print("Optimal flux:", model.get_fluxes('biomass_reaction'))
 print("Number of active reactions:", sum(abs(model.get_fluxes()) > 1e-8))
 
-# Minimize the number of reactions preserving the optimal flux
+# Transform the previous FBA optimization problem into a Sparse FBA problem (MIP).
+# Solving this problem returns the minimum number of active reactions preserving
+# the optimal flux through the biomass_reaction
 V, X = (model
         .setup(opt_tol=0.05)
         .set_fluxes_for('biomass_reaction')
@@ -34,8 +41,6 @@ V, X = (model
 print("Number of active reactions:", sum(abs(V) > 1e-8))
 ```
 
-!!! warning
-    This library is functional but still in a very early stage. API is still not stable and might be changed in future versions.
 
 ## Installation
 
@@ -75,8 +80,8 @@ MIOM includes its own lightweight and portable format (.miom) for loading and st
 To make things even easier, the method `load_gem` can import any model from this repository using the relative path, for example:
 
 ```python
-human1 = miom.mio.load_gem("@SysBioChalmers/Human-GEM.miom")
-recon3d = miom.mio.load_gem("@BiGG/Human-GEM.miom")
+human1 = miom.mio.load_gem("@SysBioChalmers/Human-GEM/v1.9.0")
+recon3d = miom.mio.load_gem("@BiGG/Recon3D.miom")
 ```
 
 Here is an example of how to load a metabolic network and maximize the flux through a target reaction using FBA, and then how to modify the original problem to implement the sparse FBA problem adding only a few lines to the original problem:
@@ -169,7 +174,7 @@ model = (miom
 
 ## Documentation
 
-The documentation of the library can be found at https://metexplore.github.io/miom/
+The documentation of the library can be found at [https://metexplore.github.io/miom/](https://metexplore.github.io/miom/)
 
 ## How to cite
 

miom/__init__.py

@@ -6,4 +6,4 @@
 )
 
 __all__ = ["load", "Solvers", "Comparator", "ExtractionMode"]
-__version__ = "0.9.0-beta.1"
+__version__ = "0.9.0-beta.2"

miom/__main__.py

@@ -2,8 +2,8 @@
 import miom
 import os
 
-def convert_list_gems(file_or_list, folder=None):
-    for input_file in file_or_list:
+def convert_list_gems(input_files, output=None):
+    for i, input_file in enumerate(input_files):
         input_file = input_file.strip()
         # Check if input_file is a valid file and it exists
         if not (miom.mio._is_url(input_file) or (os.path.isfile(input_file) and os.path.getsize(input_file) > 0)):
@@ -17,8 +17,22 @@ def convert_list_gems(file_or_list, folder=None):
             m = miom.mio.load_gem(input_file)
             print(f"Loaded network with {m.num_reactions} reactions (in-memory size: {m.object_size:.2f} MB)")
             # Concatenate folder and output file
-            if folder is not None:
-                output_file = os.path.join(folder, output_file)
+            if output is not None:
+                if isinstance(output, list) and len(output) == len(input_files):
+                    output_file = output[i]
+                    # Get absolute path of the folder of the file
+                    abspath = os.path.dirname(os.path.abspath(output_file))
+                    if not os.path.isdir(abspath):
+                        # Try to create a folder
+                        try:
+                            print(f"Creating directory {abspath}...")
+                            os.makedirs(abspath)
+                        except OSError as e:
+                            print(f"Cannot create {abspath} folder: {e}")
+                elif os.path.isdir(output):
+                    output_file = os.path.join(output, output_file)
+                else:
+                    raise ValueError("The provided output is not a folder or a list of destinations for each file")
             print(f"Exporting to {output_file}...")
             miom.mio.export_gem(m, output_file)
             # Calculate the MBs of the output_file and print it
@@ -33,12 +47,21 @@ def convert_list_gems(file_or_list, folder=None):
 def convert_gem(args):
     print(f"MIOM v{miom.__version__}")
     input = args.input
+    output = args.output
     if isinstance(input, str):
         if os.path.isfile(input):
             # Open file and get all the lines into a list
             print(f"Reading list of files from {input}...")
             with open(input, "r") as f:
-                input = f.readlines()          
+                in_files, out_files = [], []
+                input = f.readlines()
+                for line in input:
+                    if ";" in line:
+                        in_path, out_path = line.split(";")
+                        in_files.append(in_path.strip())
+                        out_files.append(out_path.strip())
+                input = in_files
+                output = out_files
         elif miom.mio._is_url(input):
             print("Imporing file from URL")
             input = [input]
@@ -50,7 +73,7 @@ def convert_gem(args):
             if not os.path.isdir(input_folder):
                 raise FileNotFoundError(f"{input_folder} is not a valid folder, a file or an URL")
             input = [os.path.join(input, f) for f in os.listdir(input)]
-    convert_list_gems(input, args.output)
+    convert_list_gems(input, output)
 
 
 def get_args():

miom/mio.py

@@ -133,10 +133,16 @@ def load_gem(model_or_path):
             matrix, the list of reactions with the lower and upper bounds, the associated
             genes and GPR rules, and the list of metabolites.
     """
+    extensions = ['.xml', '.yml', '.json', '.mat', '.miom']
     if isinstance(model_or_path, str):
         file = model_or_path
         if model_or_path.startswith("@"):
-            file = _download(DEFAULT_REPOSITORY + model_or_path[1:])
+            # Check if the file has a valid file extension
+            if not any(file.endswith(ext) for ext in extensions):
+                # Assume is a relative url pointing to a version
+                file = _download(DEFAULT_REPOSITORY + model_or_path[1:] + "/default.miom")
+            else:
+                file = _download(DEFAULT_REPOSITORY + model_or_path[1:])
         elif _is_url(model_or_path):
             file = _download(model_or_path)
         ext = pathlib.Path(file).suffix
@@ -192,15 +198,30 @@ def _cobra_to_miom(model):
         raise ImportError("Cobrapy package is not installed, "
                           "but required to read and import metabolic networks", e)
     S = create_stoichiometric_matrix(model)
-    rxn_data = [(rxn.id, rxn.lower_bound, rxn.upper_bound, rxn.subsystem, rxn.gene_reaction_rule)
-                for rxn in model.reactions]
+    subsystems = []
+    for rxn in model.reactions:
+        subsys = rxn.subsystem
+        list_subsystem_rxn = []
+        # For .mat models, the subsystem can be loaded as a string repr of a numpy array
+        if isinstance(subsys, str) and (subsys.startswith("array(") or subsys.startswith("[array(")):
+            from numpy import array
+            subsys = eval(subsys)
+            # A list containing a numpy array?
+            for s in subsys:
+                if "tolist" in dir(s):
+                    list_subsystem_rxn.extend(s.tolist())
+                else:
+                    list_subsystem_rxn.append(s)
+            if len(list_subsystem_rxn) == 1:
+                list_subsystem_rxn = list_subsystem_rxn[0]
+            subsystems.append(list_subsystem_rxn)
+        elif "tolist" in dir(rxn.subsystem):
+            subsystems.append(rxn.subsystem.tolist())
+        else:
+            subsystems.append(rxn.subsystem)
+    rxn_data = [(rxn.id, rxn.lower_bound, rxn.upper_bound, subsystem, rxn.gene_reaction_rule)
+                for rxn, subsystem in zip(model.reactions, subsystems)]
     met_data = [(met.id, met.name, met.formula) for met in model.metabolites]
-    #id_max_length = max(len(rxn.id) for rxn in model.reactions)
-    #subsyst_max_length = max((len(rxn.subsystem)) for rxn in model.reactions)
-    #metid_max_length = max((len(met.id)) for met in model.metabolites)
-    #metname_max_length = max((len(met.name)) for met in model.metabolites)
-    #metform_max_length = max((len(met.formula) if met.formula is not None else 0) for met in model.metabolites)
-    #gpr_max_length = max(len(rxn.gene_reaction_rule) for rxn in model.reactions)
     R = np.array(rxn_data, dtype=[
         ('id', 'object'),
         ('lb', 'float'),