This project is part of the Computational Biomedicine I (Fall Semester 2019) at ETH.
| Contributors |
|---|
| Deuber David |
| Adamos Solomou |
| Anagnostidis Sotiris |
| Vasilakopoulos George |
.
├── SuffixTree.cpp # implementation of Ukkonen's algorithm to build the suffix tree
├── bwt.cpp # create the index used for the alignment
├── read_genes.cpp # read from the files containing the reference and the read genes
├── randalign.cpp # implementation of the alignment strategy
├── samfile.cpp # generates the outfile in the required form
├── main.cpp # main function
└── README.md
Install tbb dependencies if not already installed (Ubuntu):
sudo apt-get install libtbb-dev
To compile:
make PARALLEL=0 # to run serial version
make PARALLEL=1 # to run parallel version
Default compiler clang. Can be changed by assigning CXX during compiling.
To run:
./<appname> index <genome_file> # index: create index, store and exit
./<appname> index_run <genome_file> <read_file_1> <read_file_2> # index_run: create index, store and run
./<appname> recover <read_file_1> <read_file_2> # recover: recover the index and run
./<appname> run <genome_file> <read_file_1> <read_file_2> #run: just run alignments
We are assuming the read_files to have the form <read_file>Cov1.fq and <read_file>Cov2.fq
The generated sam output file will be located in the same directory under the naming <read_file>Cov.generated.mod.sam
To evaluate our method we compared results generated from other read alignment tools (bowtie2), outputing in 100% of the cases the same alignments for the small genome and more than 98% of the cases the same alignments for the large genome.
There is also the option to modify the environmental varialbe PARALLEL_READS to fine tune the number of possible parallel reads, although the default value gives the best results.