mRNAOptimizer

mRNA sequence optimizer based on multi-objective genetic algorithms.

Installation

git clone https://github.com/DStarEpoch/mRNAOptimizer.git --recursive
pip install -e ".[dev]"

Optional: LinearFold (for faster RNA folding)

This package supports ViennaRNA out of the box. If you want faster folding via LinearFold, build it from the bundled submodule and add it to your PATH:

1. Clone / update the submodule

git submodule update --init --recursive

2. Build LinearFold

Linux / macOS

cd submodules/LinearFold
make

Windows (MinGW / MSYS2)

Note: mingw32-make may fail on the chmod step (which is not needed on Windows). If so, compile directly with g++:

cd submodules/LinearFold
> cd submodules/LinearFold
> g++ src/LinearFold.cpp -std=c++11 -O3 -Dlv -Dis_cube_pruning -Dis_candidate_list -o bin/linearfold_v.exe
> g++ src/LinearFold.cpp -std=c++11 -O3 -Dis_cube_pruning -Dis_candidate_list -o bin/linearfold_c.exe
> ```

The build produces two binaries:
- `submodules/LinearFold/bin/linearfold_v` (Vienna parameters)
- `submodules/LinearFold/bin/linearfold_c` (CONTRAfold parameters)

On Windows, the binaries have a `.exe` extension and a `linearfold.bat` wrapper is provided for `cdsopt` auto-detection.

#### 3. Add to PATH

**Linux / macOS**
```bash
export PATH="$(pwd)/submodules/LinearFold:$PATH"

Windows (PowerShell)

$env:PATH = "$PWD\submodules\LinearFold;$env:PATH"

Windows (CMD)

set PATH=%CD%\submodules\LinearFold;%PATH%

4. Verify

linearfold --help

If linearfold is found, cdsopt will automatically prefer it over ViennaRNA. If not, ViennaRNA is used as the fallback.

Usage

cdsopt optimize --help

Running tests

python run_tests.py -v