Merge pull request #2 from dyaus-dev/develop

Update parameter sets

.github/workflows/build.yml

@@ -3,19 +3,15 @@
 
 name: Tests
 
-on:
-  push:
-    branches: [master]
-  pull_request:
-    branches: [master]
+on: [push, pull_request]
 
 jobs:
   build:
 
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.7, 3.8, 3.9]
+        python-version: [3.7, 3.8]
 
     steps:
     - uses: actions/checkout@v2

README.md

@@ -7,32 +7,38 @@ Workflow for classifying breast cancer subtypes based on intracellular signaling
 
 ## Requirements
 
-| Language      | Dependent packages                                                                             |
-| ------------- | ---------------------------------------------------------------------------------------------- |
-| Python >= 3.7 | [biomass](https://github.com/okadalabipr/biomass), [dyaus](https://github.com/dyaus-dev/dyaus) |
-| Julia >= 1.5  | [BioMASS.jl](https://github.com/himoto/BioMASS.jl)                                             |
-| R             | [TODO] Write dependent packages here.                                                          |
+| Language      | Dependent packages                                 |
+| ------------- | -------------------------------------------------- |
+| Python >= 3.7 | [dyaus](https://github.com/dyaus-dev/dyaus)        |
+| Julia >= 1.5  | [BioMASS.jl](https://github.com/himoto/BioMASS.jl) |
+| R >= 4.0      | TCGAbiolinks, sva, biomaRt, edgeR                  |
 
-## An executable model of the ErbB network dynamics
+## Table of contents
 
-1. Use `biomass.Text2Model` to build a mechanistic model
+- [Construction](#construction-of-a-comprehensive-model-of-the-ErbB-signaling-network)
+
+- [Integration](#integration-of-tcga-and-ccle-data)
+
+- [Individualization](#individualization-of-the-mechanistic-model)
+
+- [Classification](#subtype-classification-based-on-the-ErbB-signaling-dynamics)
+
+## Construction of a comprehensive model of the ErbB signaling network
+
+1. Use `dyaus.Text2Model` to build a mechanistic model
 
    ```python
-   from biomass import Text2Model
+   from dyaus import Text2Model
 
-   Text2Model("models/erbb_network.txt").to_biomass()
+   Text2Model("models/erbb_network.txt").to_biomass_model()
    ```
 
 1. Rename **erbb_network/** to CCLE_name or TCGA_ID, e.g., **MCF7_BREAST** or **TCGA_3C_AALK_01A**
 
-1. Edit **name2idx/parameters.py**
-
-   - Add weighting factors for each gene (prefix: `"w_"`)
+1. Add weighting factors for each gene (prefix: `"w_"`) to **name2idx/parameters.py**
 
 1. Edit **set_search_param.py**
 
-   - Import `dyaus.Individualization`
-
    ```python
    import os
 
@@ -98,20 +104,20 @@ Workflow for classifying breast cancer subtypes based on intracellular signaling
        ...
    ```
 
-## Individualization of the mechanistic model
-
-### Integrate TCGA and CCLE data
+## Integration of TCGA and CCLE data
 
 [TODO] Write analysis procedure here.
 
+## Individualization of the mechanistic model
+
 ### Use time-course datasets to train kinetic constants and weighting factors
 
 1. Build a mechanistic model for parameter estimation with BioMASS.jl
 
    ```python
-   from biomass import Text2Model
+   from dyaus import Text2Model
 
-   Text2Model("models/erbb_network.txt", lang="julia").to_biomass()
+   Text2Model("models/erbb_network.txt", lang="julia").to_biomass_model()
    ```
 
 1. Add time-series data to **experimental_data.jl**
@@ -189,4 +195,4 @@ Workflow for classifying breast cancer subtypes based on intracellular signaling
 
 ## License
 
-[Apache License 2.0](LICENSE)
+[Apache License 2.0](https://github.com/dyaus-dev/dyaus/blob/master/LICENSE)

dyaus_dev/individualization.py

@@ -1,5 +1,5 @@
 from dataclasses import dataclass, field
-from typing import Dict, List
+from typing import Dict, List, Optional
 
 import pandas as pd
 
@@ -23,6 +23,9 @@ class Individualization(object):
     structure : Dict[str, List[str]]
         Pairs of proteins and their related genes.
 
+    read_csv_kwargs : dict, optional
+        Arguments to pandas.read_csv.
+
     prefix : str (default: "w_")
         Prefix of weighting factors on gene expression levels.
     """
@@ -31,10 +34,16 @@ class Individualization(object):
     species: List[str]
     tpm_values: str
     structure: Dict[str, List[str]]
+    read_csv_kwargs: Optional[dict] = field(default=None)
     prefix: str = field(default="w_", init=False)
 
     def __post_init__(self) -> None:
-        self._tpm_rle_postComBat: pd.DataFrame = pd.read_csv(self.tpm_values, index_col=2)
+        kwargs = self.read_csv_kwargs
+        if kwargs is None:
+            kwargs = {}
+        kwargs.setdefault("index_col", 2)
+        self._tpm_rle_postComBat: pd.DataFrame = pd.read_csv(self.tpm_values, **kwargs)
+        del kwargs
 
     @property
     def tpm_rle_postComBat(self):
@@ -87,6 +96,9 @@ def as_reaction_rate(
         id : str
             CCLE_ID or TCGA_ID.
 
+        x : List[float]
+            List of parameter values.
+
         param_name : str
             Parameter incorporating gene_expression_data.
 

dyaus_dev/patient_model.py

@@ -91,11 +91,11 @@ class PatientModelSimulations(InSilico):
 
     Attributes
     ----------
-    biomass_options : dict, optional
-        Arguments of biomass.run_simulation.
+    biomass_kwargs : dict, optional
+        Arguments to biomass.run_simulation.
     """
 
-    biomass_options: Optional[dict] = field(default=None)
+    biomass_kwargs: Optional[dict] = field(default=None)
 
     def _run_single_patient(self, patient: str) -> None:
         """
@@ -107,17 +107,17 @@ def _run_single_patient(self, patient: str) -> None:
             Name (ID) of each patient.
         """
 
-        options = self.biomass_options
-        if options is None:
-            options = {}
-        options.setdefault("viz_type", "average")
-        options.setdefault("show_all", False)
-        options.setdefault("stdev", True)
-        options.setdefault("save_format", "pdf")
-        options.setdefault("param_range", None)
+        kwargs = self.biomass_kwargs
+        if kwargs is None:
+            kwargs = {}
+        kwargs.setdefault("viz_type", "average")
+        kwargs.setdefault("show_all", False)
+        kwargs.setdefault("stdev", True)
+        kwargs.setdefault("save_format", "pdf")
+        kwargs.setdefault("param_range", None)
 
         biomass_model = self.import_model_package(patient.strip())
-        run_simulation(ModelObject(biomass_model.create()), **options)
+        run_simulation(ModelObject(biomass_model.create()), **kwargs)
 
     def run(self, n_proc: int = multiprocessing.cpu_count() - 1) -> None:
         """
@@ -139,11 +139,11 @@ class PatientModelAnalyses(InSilico):
 
     Attributes
     ----------
-    biomass_options : dict, optional
-        Arguments of biomass.run_analysis.
+    biomass_kwargs : dict, optional
+        Arguments to biomass.run_analysis.
     """
 
-    biomass_options: Optional[dict] = field(default=None)
+    biomass_kwargs: Optional[dict] = field(default=None)
 
     def _run_single_patient(self, patient: str) -> None:
         """
@@ -155,16 +155,16 @@ def _run_single_patient(self, patient: str) -> None:
             Name (ID) of each patient.
         """
 
-        options = self.biomass_options
-        if options is None:
-            options = {}
-        options.setdefault("target", "initial_condition")
-        options.setdefault("metric", "integral")
-        options.setdefault("style", "barplot")
-        options.setdefault("options", None)
+        kwargs = self.biomass_kwargs
+        if kwargs is None:
+            kwargs = {}
+        kwargs.setdefault("target", "initial_condition")
+        kwargs.setdefault("metric", "integral")
+        kwargs.setdefault("style", "barplot")
+        kwargs.setdefault("options", None)
 
         biomass_model = self.import_model_package(patient.strip())
-        run_analysis(ModelObject(biomass_model.create()), **options)
+        run_analysis(ModelObject(biomass_model.create()), **kwargs)
 
     def run(self, n_proc: int = multiprocessing.cpu_count() - 1) -> None:
         """

models/erbb_network.txt

@@ -288,7 +288,7 @@ E11P_Grb2_Gab1P_RasGAP_RasGTP is degraded |268|
 E11P_ShcP_Grb2_Gab1P_RasGAP_RasGTP is degraded |268|
 E11P_Grb2_Gab1P_PI3K is degraded |268|
 E11P_ShcP_Grb2_Gab1P_PI3K is degraded |268|
-ERKPP transcribes PreduspmRNA | n = 1
+ERKPP transcribes PreduspmRNA | const n = 1
 PreduspmRNA is translocated from nucleus into cytoplasm (0.22, 0.94) --> duspmRNA | kr = 0
 duspmRNA is translated into DUSP
 duspmRNA is degraded
@@ -307,7 +307,7 @@ DUSPP_ERKP is dissociated into ERK and DUSPP |301|
 ERK binds DUSPP --> DUSPP_ERK |302|
 AktP phosphorylates GSK3b --> GSK3bP || GSK3b = 1
 GSK3bP is dephosphorylated --> GSK3b
-ERKPP transcribes PrecmycmRNA, repressed by cMycP | n = 1
+ERKPP transcribes PrecmycmRNA, repressed by cMycP | const n = 1
 PrecmycmRNA is translocated from nucleus into cytoplasm (0.22, 0.94) --> cmycmRNA | kr = 0
 cmycmRNA is degraded
 cmycmRNA is translated into cMyc