Merge pull request #22 from pasmopy/develop

Update pasmopy version
pasmopy · Oct 6, 2021 · 3f96862 · 3f96862
2 parents ffd8ca5 + a23e40d
commit 3f96862
Show file tree

Hide file tree

Showing 5 changed files with 190 additions and 139 deletions.
diff --git a/README.md b/README.md
@@ -35,14 +35,42 @@ R:
 
 - [Integration of TCGA and CCLE data](#integration-of-tcga-and-ccle-data)
 
-- [Construction of a comprehensive model of the ErbB signaling network](#construction-of-a-comprehensive-model-of-the-ErbB-signaling-network)
+  - [Download TCGA clinical/subtype information](#download-tcga-clinicalsubtype-information)
+
+  - [Select samples in reference to clinical or subtype data](#select-samples-in-reference-to-clinical-or-subtype-data)
+
+  - [Download TCGA gene expression data (HTSeq-Counts)](#download-tcga-gene-expression-data-htseq-counts)
+
+  - [Download CCLE transcriptomic data](#download-ccle-transcriptomic-data)
+
+  - [Merge TCGA and CCLE data](#merge-tcga-and-ccle-data)
+
+  - [Normalize RNA-seq counts data](#normalize-rna-seq-counts-data)
+
+- [Construction of a comprehensive model of the ErbB signaling network](#construction-of-a-comprehensive-model-of-the-erbb-signaling-network)
+
+  - [From text into executable models](#from-text-into-executable-models)
+
+  - [Other tasks for incorporating gene expression levels](#other-tasks-for-incorporating-gene-expression-levels)
 
 - [Individualization of the mechanistic model](#individualization-of-the-mechanistic-model)
 
+  - [Parameter estimation](#parameter-estimation)
+
+  - [Patient-specific simulations](#patient-specific-simulations)
+
 - [Subtype classification based on the ErbB signaling dynamics](#subtype-classification-based-on-the-ErbB-signaling-dynamics)
 
+  - [Extraction of response characteristics from patient-specific simulations](#extraction-of-response-characteristics-from-patient-specific-simulations)
+
+  - [Model-based patient stratification](#model-based-patient-stratification)
+
 - [Investigation of patient-specific pathway activities](#investigation-of-patient-specific-pathway-activities)
 
+  - [Sensitivity analysis](#sensitivity-analysis)
+
+  - [Drug response data analysis](#drug-response-data-analysis)
+
 ## Integration of TCGA and CCLE data
 
 ### Download TCGA clinical/subtype information
@@ -67,7 +95,7 @@ R:
   outputSubtype("BRCA")
   ```
 
-  Output: `{TCGA Study Abbreviation}_clinic.csv` or `{TCGA Study Abbreviation}_subtype.csv`
+  Output: `<TCGA Study Abbreviation>_clinic.csv` or `<TCGA Study Abbreviation>_subtype.csv`
 
 ### Select samples in reference to clinical or subtype data
 
@@ -115,21 +143,23 @@ R:
 
     Output : `totalreadcounts.csv`
 
-### Normalization of RNA-seq counts data
+### Normalize RNA-seq counts data
 
 - Conduct noramlization of RNA-seq.
 - You can specify min and max value for truncation of total read counts.
-- If you do not want to specify values for truncation, please set `min = F` or `max = F`.
+- If you do not want to specify values for truncation, please set `min=F` or `max=F`.
 
   ```R
-  normalization(min = 40000000, max = 140000000)
+  normalization(min=40000000, max=140000000)
   ```
 
-  Output : `TPM_RLE_postComBat_{TCGA}_{CCLE}.csv`
+  Output : `TPM_RLE_postComBat_<TCGA>_<CCLE>.csv`
 
 ## Construction of a comprehensive model of the ErbB signaling network
 
-1. Use `pasmopy.Text2Model` to build a mechanistic model
+### From text into executable models
+
+1. Use [`pasmopy.Text2Model`](https://pasmopy.readthedocs.io/en/latest/model_development.html) to build a mechanistic model
 
    ```python
    import os
@@ -140,11 +170,24 @@ R:
    Text2Model(os.path.join("models", "erbb_network.txt")).convert()
    ```
 
+1. Rename `erbb_network/` to CCLE_name or TCGA_ID, e.g., `MCF7_BREAST` or `TCGA_3C_AALK_01A`
+
+   ```python
+   import shutil
+
+   shutil.move(
+       os.path.join("models", "erbb_network"),
+       os.path.join("models", "breast", "TCGA_3C_AALK_01A")
+   )
+   ```
+
+### Other tasks for incorporating gene expression levels
+
 1. Add weighting factors for each gene (prefix: `"w_"`) to [`name2idx/parameters.py`](models/breast/TCGA_3C_AALK_01A/name2idx/parameters.py)
 
    ```python
+   from pasmopy import Model
    from pasmopy.preprocessing import WeightingFactors
-   from biomass import Model
 
    from models import erbb_network
 
@@ -175,21 +218,10 @@ R:
    }
 
    weighting_factors = WeightingFactors(model, gene_expression)
-   weighting_factors.add()
+   weighting_factors.add_to_params()
    weighting_factors.set_search_bounds()
    ```
 
-1. Rename `erbb_network/` to CCLE_name or TCGA_ID, e.g., `MCF7_BREAST` or `TCGA_3C_AALK_01A`
-
-   ```python
-   import shutil
-
-   shutil.move(
-       os.path.join("models", "erbb_network"),
-       os.path.join("models", "breast", "TCGA_3C_AALK_01A")
-   )
-   ```
-
 1. Edit [`set_search_param.py`](models/breast/TCGA_3C_AALK_01A/set_search_param.py)
 
    ```python
@@ -261,7 +293,9 @@ R:
 
 ## Individualization of the mechanistic model
 
-### Use time-course datasets to train kinetic constants and weighting factors
+### Parameter estimation
+
+Here, we use phospho-protein time-course datasets to train kinetic constants and weighting factors.
 
 1. Build a mechanistic model to identify model parameters
 
@@ -312,10 +346,10 @@ R:
    breast_cancer_models = []
    path_to_models = os.path.join("models", "breast")
    for model in os.listdir(path_to_models):
-        if os.path.isdir(os.path.join(path_to_models, model)) and (
-            model.startswith("TCGA_") or model.endswith("_BREAST")
-        ):
-            breast_cancer_models.append(model)
+       if os.path.isdir(os.path.join(path_to_models, model)) and (
+           model.startswith("TCGA_") or model.endswith("_BREAST")
+       ):
+           breast_cancer_models.append(model)
    # Set optimized parameters
    for model in breast_cancer_models:
        shutil.copytree(
@@ -324,7 +358,7 @@ R:
        )
    ```
 
-### Execute patient-specific models
+### Patient-specific simulations
 
 - Use `pasmopy.PatientModelSimulations`
 
@@ -353,26 +387,30 @@ R:
 
 ## Subtype classification based on the ErbB signaling dynamics
 
-1. Extract response characteristics from patient-specific simulations
+### Extraction of response characteristics from patient-specific simulations
 
-   ```python
-   simulations.subtyping(
-       fname=None,
-       dynamical_features={
-           "Phosphorylated_Akt": {"EGF": ["max"], "HRG": ["max"]},
-           "Phosphorylated_ERK": {"EGF": ["max"], "HRG": ["max"]},
-           "Phosphorylated_c-Myc": {"EGF": ["max"], "HRG": ["max"]},
-       }
-   )
-   ```
+- Execute `subtyping()`
+
+  ```python
+  simulations.subtyping(
+      fname=None,
+      dynamical_features={
+          "Phosphorylated_Akt": {"EGF": ["max"], "HRG": ["max"]},
+          "Phosphorylated_ERK": {"EGF": ["max"], "HRG": ["max"]},
+          "Phosphorylated_c-Myc": {"EGF": ["max"], "HRG": ["max"]},
+      }
+  )
+  ```
 
-1. Visualize patient classification by executing [`brca_heatmap.R`](classification/brca_heatmap.R)
+### Model-based patient stratification
 
-   ```bash
-   $ cd classification
-   # $ Rscript brca_heatmap.R [n_cluster: int] [figsize: tuple]
-   $ Rscript brca_heatmap.R 6 8,5
-   ```
+- Run [`brca_heatmap.R`](classification/brca_heatmap.R)
+
+  ```bash
+  $ cd classification
+  # $ Rscript brca_heatmap.R [n_cluster: int] [figsize: tuple]
+  $ Rscript brca_heatmap.R 6 8,5
+  ```
 
 ## Investigation of patient-specific pathway activities
 

diff --git a/drug_response/README.md b/drug_response/README.md
@@ -1,6 +1,6 @@
-# CCLE drug response data analysis and visualization
+## CCLE drug response data analysis and visualization
 
-## Drug response data
+### Drug response data
 
 #### Description:
 
@@ -14,43 +14,49 @@
 
 - Barretina, J., Caponigro, G., Stransky, N. _et al_. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. _Nature_ **483**, 603–607 (2012). https://doi.org/10.1038/nature11003
 
-## Usage
-1. Prepare sample.txt and gene.txt, which contain the names of the samples and genes, respectively.
+### Usage
+
+1. Prepare [`sample.txt`](data/sample.txt) and [`gene.txt`](data/gene.txt), which contain the names of the samples and genes, respectively.
+
 1. Open R
 
    ```bash
    $ R
-   ```  
+   ```
 
-1. Load CCLE_normalization.R
+1. Load `CCLE_normalization.R`
 
    ```R
    source("calc_erbb_ratio.R")
    ```
 
-1. Read sample.txt and gene.txt  
-If you want data for a specific samples or genes, you will need to create a list of those samples as `sample.txt` or `gene.txt` and read it here. 
+1. Read [`sample.txt`](data/sample.txt) and [`gene.txt`](data/gene.txt)
+
+   If you want data for a specific samples or genes, you will need to create a list of those samples as [`sample.txt`](data/sample.txt) or [`gene.txt`](data/gene.txt) and read it here.
+
    ```R
    gene <- scan("gene.txt", what="character")
    #sample <- scan("sample.txt", what="character")
    ```
 
 1. Create TPM/RLE normalized RNA-seq data matrix of selected samples and genes
+
    ```R
    CCLEnormalization(gene, sample = NULL)
    ```
-   Output: CCLE_normalized.csv (TPM/RLE normalized RNA-seq data for specific samples or genes)
 
-1. Calculate receotor ratio  
-Calculate EGFR/(ERBB2+ERBB3+ERBB4) using CCLE_normalized.csv.  
-If you want to run this coad, you need to prepare gene.txt containing the names of these 4 genes.
+   Output: [`CCLE_normalized.csv`](data/CCLE_normalized.csv) (TPM/RLE normalized RNA-seq data for specific samples or genes)
+
+1. Calculate EGFR/(ERBB2+ERBB3+ERBB4) using [`CCLE_normalized.csv`](data/CCLE_normalized.csv)
+
+   If you want to run this code, you need to prepare `gene.txt` containing the names of these 4 genes.
+
    ```R
    CCLE_normalized <- read.csv("CCLE_normalized.csv", row.names = 1)
    receptor_ratio(data = CCLE_normalized, num = 30)
    ```
-   Output: "ErbB_expression_ratio.csv"
 
-
+   Output: [`ErbB_expression_ratio.csv`](data/ErbB_expression_ratio.csv)
 
 1. Drug response analysis and visualization
 

diff --git a/requirements.txt b/requirements.txt
@@ -1,6 +1,6 @@
 pasmopy
 biomass==0.5.2
-matplotlib
+matplotlib>=3.0
 numpy>=1.17
 pandas>=0.23
 seaborn>=0.11

diff --git a/tests/test_patient_specific_modeling.py b/tests/test_patient_specific_modeling.py
@@ -5,8 +5,7 @@
 from typing import Callable, List
 
 import numpy as np
-from biomass import Model
-from pasmopy import PatientModelSimulations, Text2Model
+from pasmopy import Model, PatientModelSimulations, Text2Model
 from pasmopy.preprocessing import WeightingFactors
 
 from .C_erbb_network import *
@@ -68,7 +67,7 @@ def test_model_construction():
     }
 
     weighting_factors = WeightingFactors(model, gene_expression)
-    weighting_factors.add()
+    weighting_factors.add_to_params()
     weighting_factors.set_search_bounds()
     # Edit observable.py, update normalization
     with open(