SD2E / CDM

Mirror of SD2E CDM

README Last Updated by Mohammed on 4/9/21

CDM

Installation

1. Clone this repository into the environment of your choice (directory, conda env, virtualenv, etc). Conda envs are recommended. *** Ignore steps 2 and 3 if you want more flexibility to install versions of your own choice
2. Using command-line or terminal, Navigate inside the cdm directory. You should be at the same level of requirements.txt.
3. Run pip3 install -r requirements.txt .
4. Using command-line or terminal, navigate to the directory in which you cloned this repo (not inside the cdm directory itself). This should be 1 level higher than where you were in the previous step.
5. Run pip3 install ./cdm or pip3 install -e cdm . This will install the cdm package and make it visible to all other repositories/projects you have in the current environment. The -e option stands for "editable". This will install the package in a way where any local changes to the package will automatically be reflected in your environment. See this link for more details.

Note 1: Do not do pip3 install cdm! Because that will install a different cdm package that exists on pypi. Instead do pip3 install ./cdm so pip knows to look at for a directory.

• For editable mode, it doesn't matter so feel free to do pip3 install -e cdm.

Note 2: On TACC you might not be able to install this package unless you use -e and/or --user: e.g. pip3 install -e cdm --user.

Running an Example Notebook

There are two example notebooks in the example directory with associated data. Navigate to the "example" directory and run example_notebook.ipynb is for the B. subtilis examples and ecoli_example_notebook.ipynb is for the E. coli example.

Data

Datasets included are:

• E. coli -- differential expression analysis for all conditions (ecoli_additive_design_df.csv) and a local version of the original EcoliNet (CX.INT.EcoliNet.v1.4039gene.67494link.txt) as well as a version where locus tags are translated to gene symbols (CX.INT.EcoliNet.v1_translated.csv). Use the translated version to join with the differential expression analysis data. For details on EcoliNet, visit: https://www.inetbio.org/ecolinet/

• B. subtilis -- two files for differential expression analysis, training/validation data: (additive_design_df_1.csv) and test data collected from experiments after the model was trained (additive_design_df_2.csv). The network used for B. subtilis can be found here: (bacillus_net.csv) with details here: https://www.embopress.org/doi/full/10.15252/msb.20156236

Authors

• Mohammed Eslami, Netrias LLC
• Hamed Eramian, Netrias LLC