A C++14 library and software to extract unique sequences from FASTA alignments. These unique sequences can be found by scanning the entire alignment, providing a position and window length, or providing a query sequence.
Building the library and binaries requires a C++14 compiler. The build process requires cmake version 3.11 or later. Sequence query uses Smith-Waterman alignment, implemented here. This code is included as a submodule.
The repository comes with submodules, so to clone use
git clone --recurse-submodules https://github.com/tonymugen/analyzeAlignmentsNext, create a build directory
cd analyzeAlignments
mkdir buildFinally, run cmake to build and install the software
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
cmake --build .
cmake --install .Installation may require root privileges.
Optionally, one can also build the unit tests. These require Catch2, although its installation is taken care of by cmake. To build the tests, create a build-Tests directory, say, and run
cd build
cmake -DCMAKE_BUILD_TYPE=Test -DBUILD_TESTS=ON ..
cmake --build .To run the tests from the build directory, simply run
./testsTwo binaries are built as part of the project. Command line flags and their descriptions can be printed by running the programs without parameters.
The homoruns binary takes an alignment and sliding window parameters (window and step size) and outputs unique sequence counts for each window. Sequences themselves are not saved, but counts are reported for each unique sequence.
The extractWindow binary takes an alignment and either a start window position and length or a query sequence. It returns all unique sequences in the window (or best matches to the query) with their counts. The sequences can be optionally sorted by their counts in descending order.