Simply download the executable from the latest release, via the following
wget https://github.com/GZHoffie/bbs/releases/download/v0.2.0/bbsAlternatively, build skiver from the source code. Install rust, and build using
https://github.com/GZHoffie/bbs.git
cd bbs
# If default rust install directory is ~/.cargo
cargo install --path . --root ~/.cargo
BBS is available by running
bbs
Currently, BBS supports 2 input formats,
dna_storage_toolkitformat.microsoftformat: similar to the CNR dataset. Below is a quick demo.
# Download the Microsoft CNR dataset
git clone https://github.com/microsoft/clustered-nanopore-reads-dataset.git
# Assume that we are still in the build/ directory
# To output the reconstructed sequence directly, use `./bbs -i <input_clusters> -l <read_length> > <output_file_name>`
bbs clustered-nanopore-reads-dataset/Clusters.txt -l 110 > output.txtFor more detailed output, use the following command.
bbs clustered-nanopore-reads-dataset/Clusters.txt -l 110 -o output_verbose.csvIn the output csv file, there will be 5 fields, indicating the index of the cluster, reconstructed sequence, the value of k used, the total path weight, and the confidence value.
read_id,reconstruction_result,k,path_weight,confidence
1,ACCATAATGCGTGGGGCCGACCTCGGAATGCGGTCTCCATGCGCGTTTCCTCCAACCTAAGGTAGCCTGTAGTTCATTGGACCTCTGATGGCGCTTATAGAAACCGGGAA,11,-14.9066,0.909951
2,TCGAAGCAGTAGGGCCTACCAAATAGGTTGGTCCTCCGTTGTATCTAAGGATTGAGTTTACCTGGCTTACACGGCAGGTACCGCCAATCTCGTCCGGCTCCGCGGCATCC,8,-32.2539,0.950223
3,AGTTAACGTCCCACGGCGAGGCACTCTTGATCCCCACCTTCAAGAGGTGTACCGGATCATGGAGAACAAGCATACGTCGCACGCACACCATTGGACGGCGAGTGCCGAGT,10,-44.4446,0.853414
Use the following for a detailed guide on other input parameters.
bbs -hGu Z, Xin H, Sharma P, Goh GY, Wong L, Nagarajan N. Efficient trace reconstruction in DNA storage systems using bidirectional beam search. iScience. 2025 Oct 21;28(11):113791. doi: 10.1016/j.isci.2025.113791. PMID: 41280695; PMCID: PMC12630026.